COURT FILE NO.: CV-14-00502628-0000

DATE: 20240419

ONTARIO

SUPERIOR COURT OF JUSTICE

BETWEEN:

GARY CURTIS

Plaintiff

– and –

THE BANK OF NOVA SCOTIA

Defendant

Self-Represented

Bonnie Roberts Jones and Rayaz Khan, for the Defendant

HEARD: February 26, 27, 28, 29, March 1, 4, and 5, 2024

AKAZAKI J.

REASONS FOR JUDGMENT

Table of Contents

INTRODUCTION AND CONDUCT OF THE TRIAL.. 2

ISSUE #1 – LIBEL.. 6

ISSUE #2 – JUSTIFICATION.. 11

ISSUE #3 – QUALIFIED PRIVILEGE.. 20

ISSUE #4 – THE MALICE EXCEPTION TO QUALIFIED PRIVILEGE.. 25

ISSUE #5 – DAMAGES. 29

General Damages and Aggravated Damages. 30

Special Damages. 31

Punitive Damages. 35

PREJUDGMENT INTEREST AND TOTAL DAMAGES. 37

COSTS. 37

INTRODUCTION AND CONDUCT OF THE TRIAL

[1] On April 25, 2012, the professional life of Gary Curtis was about to be turned on end. Until that day, he had served twelve years as a Mortgage Development Manager, or MDM, at the Bank of Nova Scotia. As a group, MDM’s had “sold” over $1 billion in residential mortgage lending in the Toronto region in 2011. Curtis had been one the Toronto Region’s top producers for some time. There was some dispute whether he was #5 or #4 in overall performance. He contended the bank incorrectly tallied his sales below his group leader’s to keep him out of an elite group entitled to special perks.

[2] Curtis now found himself face to face with a mortgage fraud investigator. She showed him various mortgage application files in which his customers appeared to have submitted fraudulent income and employer documentation. One or more of the sales referrals were realtors who, it turned out, were also licensed as mortgage brokers. MDM’s were not permitted to accept mortgage broker referrals because another section of the banks’ mortgage department handled them. He was, however, allowed to accept referrals from realtors who also had mortgage broker licences. Much of the investigation and the trial picked over the niceties of Curtis’ adherence to this policy.

[3] Curtis found himself unable to account for having allowed the bank to process these applications. The bank suspended him with pay, pending further investigation of “procedural irregularities.” Since he was on 100% commission, his employer was unclear how it was any different from suspension without pay. Fearing that he would be left without income for an indefinite period, he submitted his resignation and looked for work at the bank’s competitors. This was a blunder. He disentitled himself from pleading unfairness. He made it easier for the bank to conclude he left to escape or to “outrun” a fraud investigation. Perhaps if the investigator had made it clearer that he was the subject of the fraud investigation, he could have reconsidered the wisdom of quitting.

[4] The bank accepted his resignation and closed his employment file with the code “ineligible for rehire.” On May 3, 2012, the investigator initiated a “SIFT” alert on the Bank Crime Prevention and Investigation Office (BCPIO) of the Canadian Bankers Association (CBA) database based on the following grounds in the SIFT Database Input Request Form:

Strong evidence was found which establishes that Curtis knowingly uttered fraudulent documents on several applications. Said documents included falsified NOA's [Notices of Assessment], letters of employment, pay stubs, and financial statements as well as Bank documents containing forged customer signatures. These acts taken by themselves form a prima facie case in relation to the alleged fraud. Curtis resigned during the course of the investigation.

[5] Curtis received offers of employment from the mortgage departments of two other major banks. As a result of the SIFT alert, each offer was soon withdrawn after the prospective employers contacted the defendant and the defendant “Advised that subject had submitted fraudulent documentation in support of mortgage applications.” The third bank rejected his application after having received that information from the defendant. The defendant maintained the SIFT alert for seven years and effectively stopped Curtis from being employed by a major financial services institution when he was reaching the pinnacle of his career.

[6] The above synopsis should make the nature of the case, and the issues in it, rather obvious. To a self-represented litigant whose journey took him through the internet’s marketplace of legal jargon, he cannot be faulted for having laden his case with shiny objects he picked up along the way. For the benefit of the plaintiff, I must explain why the expenditure of so much pre-trial effort and trial time on outwitting the bank’s lawyers detracted from his ability to prove his damages.

[7] This action started as a suit for wrongful or constructive dismissal and defamation. The employment law portion was stayed in 2015 pending judicial review of a Canada Labour Code determination. After that judicial review was dismissed, the plaintiff withdrew the employment law claims and amended the claim reducing the case to one founded in defamation. This was a stroke of genius, especially for an autodidact. The employment law case was never going to succeed. He had resigned and was not fired or even constructively dismissed. He then reversed course and sought leave for further amendments to plead various other private law theories, some of them back-door attempts to reintroduce the employment law case.

[8] On December 17, 2019, Associate Justice Wiebe denied leave for some amendments and allowed some. The decision is reported at 2019 ONSC 7539. He disallowed new pleadings for bad faith breach of the employment contract, negligent or bad faith investigation, intentional interference with contractual relations, and inducing breach of employment contract. He allowed, with some reluctance, new pleadings for bad faith and malicious referencing “as an adjunct to … defamation,” intentional interference with economic relations, breach of customer contract and breach of privacy, and intentional infliction of emotional distress.

[9] The case took a decade to reach trial. There were many interlocutory skirmishes and proceedings. Each side sought declarations against the other to be named as vexatious litigants. Ironically, this was to be a summary trial. The plaintiff obtained the trial date at a February 21, 2023, case conference before Chalmers J. in which the plaintiff represented to the court that “the matter is straightforward and the only matter remaining in issue is the assessment of damages.” This could have been another flash of brilliance. He could have devoted all his remaining energy into preparing his damages case. Unfortunately, he soon distracted himself. His damages case likely had considerable potential, but it suffered from lack of convincing evidence. He spent most of the trial pursuing unnecessary tangents.

[10] At the opening of trial, the plaintiff brought motions that detracted from the strength of his case by making his case appear disorganized and frivolous. First, he sought further amendments to his pleadings. He sought to lower or to refine the quantum of damages sought under various headings and added “breach of duty of honest performance” and “negligence” as labels for legal theories. The defendant did not oppose the amendment of the damage’s figures. I allowed the amendment for the duty of honest performance since it seemed to be subsumed into the issues in the defamation case. I did not allow the negligence claim because the pleaded facts did not support it.

[11] The plaintiff then brought a motion to strike the allegations of justification based on fraud allegations. He argued the defendant was estopped, res judicata, from contesting the finding of a labour adjudicator that there was no evidence of fraud in support of the bank’s claim for retention of unpaid commissions. I denied the motion on the basis that the first element of res judicata, identity of issue, was not made out: Velocity Standardbreds v. Tackoor, 2019 ONSC 1995, at para. 12. In my view, the unpaid commissions and defamation claims involve entirely different legal and factual matrices. Moreover, the underlying rationale for the doctrine, expressed in terms of finality, multiple jeopardy, judicial economy, and abuse of process, cannot be met if I must sift through the evidence to determine which fraud issues were before the labour adjudicator and which are germane to the defamation case. See Collins and Hanson v XYTEX Corporation, et al, 2017 ONSC 3341, at paras. 23-25.

[12] The plaintiff did not have the benefit of legal training or the intuition of a barrister to know that the bank’s defence of justification, in the absence of the ability to prove it, only drew attention to the weakness of the qualified privilege defence. The absence of proof of his involvement in the mortgage frauds made it difficult for the bank to seek refuge in the necessity to publish the defamation and to justify having kept his name on the SIFT registry for seven years.

[13] Many times, during the trial, the plaintiff’s interventions including those from the witness box called on me verbally to corral him into presenting his case in a more focused and linear manner. As the Court of Appeal recently reiterated in R. v. Bancroft, 2024 ONCA 121, at paras. 13-15, the trial judge must assist self-represented parties to ensure their case emerges with “full force and effect” without “descending into the arena and losing neutrality.” The plaintiff had come to court to do battle so much on extraneous issues that he did not heed his own advice to Justice Chalmers that the case was a straightforward libel case in which the trial was needed to present evidence of his damages.

[14] Even when the bank’s counsel conceded in her opening that the plaintiff had made out a prima facie case in defamation, the plaintiff continued to pursue theoretical torts and tried to shoehorn marginal facts into them. The practice of repackaging goods with new wrapping and labels in the hope that the buyer will not see through it is not limited to self-represented plaintiffs. As Oliver Wendell Holmes Jr. famously remarked about his time as a judge, “The life of the law has not been logic: it has been experience.” The facts, and only the facts, drive the legal theory and the remedy. Nor is the life of the law based on incantations such as those taught at Hogwarts. The labels in legal textbooks categorize the types of facts that common law courts have considered as justifying remedies. They are not words, spoken in English or Latin, which bring down thunderbolts from the sky.

[15] When the action is founded on written words communicated by the defendant to third parties and the damages arise from their falsity, libel is the causes of action. In other instances, where the words are not impugned for their defamatory meaning but are acts of saying things or planning to do things that are then done, some other torts might be in play. It is in those instances where torts such as conspiracy and economic interference could describe unlawful acts warranting civil remedies. The fact that Associate Justice Wiebe allowed the plaintiff to make various amendments does not mean I have to spill pints of ink explaining why the labels do not fit the evidence at this trial.

[16] The case, on the entirety of the evidence before the court, was about the bank’s communication to Curtis’ prospective employers that he had submitted fraudulent mortgage application documents, with the clear inference that he had knowingly done so or had even created the fraudulent documents. The defendant conceded the elements of defamation, namely the publication of a statement (a) referring to the plaintiff, (b) to a third party, and (c) conveying defamatory meaning or innuendo in the sense of lowering the plaintiff’s reputation in the estimation of reasonable persons. See Grant v. Torstar Corp., 2009 SCC 61, [2009] 3 SCR 640, at paras. 28ff. Therefore, the court must find the defendant liable unless the defendant was successful in its positive defences.

[17] The bank defended the case by asserting two such defences:

1. justification based on the truth of the statements, and

2. qualified privilege based on the occasion of a Canadian Bankers Association facility for fraud prevention, for the purpose of public protection and on the occasion of follow-up communication to other banks, akin to employment references.

[18] The law does not strictly compartmentalize these defences. Both the cause of action and defences are concentric in that the common law has developed rules for balancing the traditional protection of damage to reputation and emergent liberal principles of public interest and freedom of expression. Justification requires a de novo inquiry into the evidence to rebut the presumption that a defamatory statement was false. In contrast, qualified privilege and the malice exception require an ex ante exploration of the defendant’s compulsion or duty to speak to raise reasonable and necessary concerns and the sufficiency of evidence prior to the making of a communication that turns out to be untrue. Thus, the law of defamation puts the onus on the publisher of the harmful statement to ensure it is true. It also recognizes that in certain occasions a party needs to publish information that may turn out to be false.

[19] The plaintiff’s response to the justification defence was to highlight the evidence that the defamatory words were unfounded. His response to the qualified privilege placed an enormous emphasis on malice. He almost backed into the response that the publication was unwarranted. The defendant failed to present any admissible evidence to prove truth as a defence to the libel claim. Rather, much of the defence’s case was an attempt to impose on Curtis as an MDM the primary responsibility of preventing fraudulent mortgage applications from getting to the underwriting stage. This undermined the qualified privilege defence because the BCPIO criteria for reporting an employee to SIFT excluded lack of due diligence or other competence-related failing. Although the plaintiff failed to establish the report to SIFT was motivated by malice, the malice exception eroded the qualified privilege once the plaintiff sued and the bank kept him on the SIFT register for the full seven years without any review of the soundness of the original investigation findings. Both defences therefore failed to oust the bank’s liability to the plaintiff for having defamed him.

[20] Although some harm is presumed in defamation cases without formal proof, the plaintiff presented a claim for many millions of dollars in economic damages. This part of the case suffered from lack of evidence. Therefore, despite an outcome that will take funds from the pocket of the defendant to pay the plaintiff, the plaintiff failed to achieve the success in his damages claim that he had hoped.

ISSUE #1 – LIBEL

[21] I will deal with the plaintiff’s libel case even though the bank conceded he had a prima facie case, because the content and circumstances of the defamatory communication are relevant to the defences. There was also a factual dispute on one point, regarding the plaintiff’s status as “not eligible for rehire.” For reasons I will outline, that designation was not defamatory as an internal statement, and there was insufficient evidence of the bank having communicated it outside its own organization.

[22] As mentioned at the outset, the plaintiff was a high-performing MDM with the bank at the time of the incidents in question. Prior to that, he had been employed for about six years as a retail branch teller and personal banking officer, including a period when he obtained his degree in administrative studies from York University. He also completed various banking and investment courses with the Institute of Canadian Bankers and the Canadian Securities Institute. He received numerous commendations. The witness from the bank who was the plaintiff’s ultimate superior, Vice-President Kevin Conroy, admitted that Curtis had been a top performer who had been invited on several junkets organized by the bank. Indeed, the incentive to keep him and other MDM’s who had been under investigation meant the bank could not have wanted to push him out.

[23] The plaintiff contended that the mortgage delinquency rate of is mortgage portfolio was much lower than other MDM’s, including that of his Regional Manager. The plaintiff repeatedly pointed to the bank’s failure to produce his personnel and performance records as proof of its intention to hide malice. He also showed in the documents that the bank had made a mistake on his last performance ranking report to deny him a major annual perk. If his performance was better than others, including his manager’s, he argued, jealousy was a strong incentive to investigate his files and push him out. The plaintiff has had twelve years to consider what had transpired and rationalize the events based on ill will toward him. I will deal with the consequences of the litigation experience on the plaintiff when I deal with his damages.

[24] The lack of evidence made it hard to attach weight to Curtis’ subjective opinions regarding the reasons for the bank suddenly driving out a star employee. As I discussed with the parties during the closing submissions, I also heard no evidence that the bank racially discriminated against Curtis, even though I was aware of the potential role of unconscious bias in the bank employees’ conclusion that he had attempted to “outrun” the fraud investigation. That word, employed by Conroy, certainly has negative connotations rooted in the history of North American race relations.

[25] The only evidence of what prompted the investigation was contained in the written reports disclosed by the defendant. During an unrelated investigation, the Security and Investigation (S&I) unit of the bank came upon potentially fraudulent employment and income documentation attached to mortgage applications processed by Curtis. This resulted in a closer examination of 18 files. The investigator identified 16 which contained suspicious documents, two of which were declined applications. That left 14 mortgages on the bank’s books that may have been obtained by a phenomenon that Conroy labelled “shelter fraud.” Shelter fraud refers to residential financing obtained by inflating income or falsifying employment details. The estimated risk to the bank was $4.7 million, although none of the mortgages were in default.

[26] I should pause to observe that the nature of the risk was a complex one, and that neither side dealt with it in much detail in their presentation of evidence. The bank, as will be seen in the next section, contended there was a public interest in protecting against mortgage fraud. In reality, the main consequence of shelter fraud was risk to the bank. This included manipulating the purchase price to inflate the relative value of deposits, in order to elude the requirement for CMHC mortgage insurance for low-deposit home purchases. Curtis repeated at length that since none of the mortgages were in default, that was evidence that many of the suspect files did not involve fraud. He also contended that because one of the applications had been for a bank employee in the risk division, it was hard to imagine that fraud could be involved. I found his evidence and his submissions on these points to have demonstrated complete lack of understanding of the type of fraud in question. In an indirect fashion, his lack of comprehension of the issue was evidence that he could not have been involved.

[27] In February 2012, Senior S&I Investigator Jessica Feiereisen (“the investigator”) telephoned Conroy to let him know that one of his MDMs, the plaintiff, was under investigation. The plaintiff questioned why Conroy did not alert him, and Conroy testified that to do so could compromise the investigation. I agree with Conroy’s explanation and consider the plaintiff’s complaint baseless.

[28] On April 25, 2012, the plaintiff had been scheduled to attend a meeting of the Toronto Region MDM’s. However, the day before, he was summoned to attend a meeting at the BNS headquarters at Bay and King the next day.

[29] On April 25, 2012, Senior S&I Investigator Jessica Feiereisen (“the investigator”) and Employee Relations Manager Shirley Roberts, as well as a junior member of S&I, met with Curtis. Roberts took typewritten notes of the interview, which appeared to explain why the investigator did not take notes and why the latter’s report ended with input to April 24 and did not include the interview. The investigator did not testify. Roberts refreshed her memory by reference to the notes.

[30] According to Roberts’ report to the Employment Relations department and her testimony, the interview covered the following topics, and she noted her impressions after discussion of each segment:

1. Curtis had dealings on mortgage applications submitted by clients of one Mr. O, a dual licensee realtor and mortgage broker. The investigator showed him documentation from which Curtis knew or ought to have known that Mr. O was a mortgage broker.

2. Curtis handled a mortgage application submitted by Mr. O for a personal loan. Curtis gave confused or conflicting information about his knowledge of Mr. O’s qualifications as a realtor and employment status in the previous two to four years. According to Roberts, Curtis changed his explanation regarding the length of time he knew Mr. O and concluded that Curtis’s responses “were not credible and it is our conclusion that Mr. Curtis knowingly submitted fraudulent documentation for Mr. O’s personal finance mortgage.” Curtis was told his responses were not credible and was given an opportunity to amend his response; but he did not.

3. After the interview, Roberts found out that Mr. O had been blocked by a separate division of the mortgage department.

4. Curtis had dealings with another dual licensee, one Mr. Roberts found Curtis’ statement that some realtors submit mortgage applications was not credible. She concluded that Curtis had accepted and submitted fraudulent and forged documentation and knowingly breached bank policies regarding the sources of mortgage documentation.

5. Curtis maintained that he met with customers, but he was shown documents illustrating that he sometimes dealt with them by email. Roberts concluded that he “had not been truthful during the interview.” He was given a chance to change his answers, but he remained adamant that he met with all customers.

6. Curtis stated that he would perform internet searches for employment verification to see if there companies that matched the employers on mortgage forms, but “he indicated that it was not his role.” It was the role of another unit to verify the documents. Several documents with “red flags” were shown to him. Roberts concluded that Curtis knowingly submitted fraudulent documentation or turned a blind eye.

7. Curtis was shown a mortgage deal that had also been submitted to another unit at the bank. The two sets of forms had inconsistent employment information. He was asked how he could proceed with his deal. Roberts concluded that his response, that he could not tell that the information was fraudulent or incorrect, lacked judgment and integrity.

8. Curtis was questioned about $61,800 he had paid to his girlfriend, a person who may have been training to become a mortgage broker. He maintained that he was not receiving mortgage deals from her. Roberts found him to have been evasive and not credible.

[31] Many of the conclusory comments were expressed as: “we concluded.” Since the investigator did not testify, the only thing that turns on this is that the court only had the benefit of Roberts’ opinion and that the collective “we” possibly meant to introduce inadmissible hearsay. Objectively, the defence presented a factual narrative premised on the fraudulent mortgage documentation. Apart from the investigator’s notes regarding a cold call and the discovery that the mailing address was a commercial mailbox, which I will discuss in the section regarding the defence of justification, the only evidence of a potential fraudulent document was a CRA notice of assessment that contained a line for Ontario non-refundable tax credits with a bold font. The court has no expertise in document examination at that level of subtlety. Moreover, I did not see the relevance of a tax credit to an allegation of falsified income. The bank could have submitted the document to a document examiner to show it had been altered when originally submitted.

[32] At this juncture, the relevance of the absence of proof of fraudulent documents or forgeries, let alone Curtis’ knowledge of them, is the absence of factual foundation of the published defamation, to which I will turn in a moment. For now, I pause to observe that, in the absence of proof of the frauds or forgeries, the conclusion in Roberts’ report that Curtis knowingly accepted them for processing by the bank’s underwriters and legal personnel lacked a factual bridge from suspicion to conclusion. There was no proof of any fraud on which the court could rely, let alone of Curtis’ involvement. Moreover, all of the evidence of the circumstances of the interview indicated that Curtis faced discipline or termination for careless or breach of bank policies, and that he remained unaware that he was the target of the fraud investigation. Being duped by fraudsters and having participated in fraud are not the same.

[33] The same day as the interview, the plaintiff received a letter from Conroy on Roberts’ letterhead stating that he was under investigation for “procedural irregularities” and that he was suspended with pay. It also stated:

Depending on the outcome of this investigation, you could be subject to disciplinary action up to and including termination of your employment. We will advise you as to your employment status as soon as the investigation is complete.

[34] The plaintiff made inquiries to determine how he would be compensated and there was no satisfactory answer, given that he was a commissioned salesperson. It was conceded by the defence that the $40,000 annual stipend stated on his original employment contract only applied to the first year. Fearing that an investigation could take an indeterminate length of time, he submitted his resignation on April 30 and started immediately to look for employment at the other major banks headquartered in Toronto, BMO, CIBC and RBC.

[35] On May 2, 2012, the bank accepted the resignation in a letter stating:

We would like to acknowledge receipt of your letter of resignation effective April 30, 2012, and we accept your resignation effective April 30, 2012.

Further to our suspension letter of April 25, 2012, we have now completed our investigation into irregular practices, and it is our conclusion that you breached the Guidelines for Business Conduct during the course of your employment with the Bank.

[36] Any reasonable person reading this letter would have concluded that Curtis’ employment offence was dealing with mortgage brokers against bank guidelines.

[37] On May 3, 2012, Roberts circulated her “Final Irregular Practice Report,” the contents of which I have already summarized as reflecting the course of the interview.

[38] On May 3, 2012, the investigator submitted a SIFT Database Input Request Form. She copied and pasted in its entirety Roberts’ report and added the following opening gloss, which bears repeating:

Strong evidence was found which establishes that Curtis knowingly uttered fraudulent documents on several applications. Said documents included falsified NOA’s letters of employment pay stubs and financial statements as well as Bank documents containing forged customer signatures. These acts taken by themselves form a prima facie case in relation to the alleged fraud. Curtis resigned during the course of the investigation.

[39] The meaning of the word “uttered” was the topic of some discussion at trial. It is an antiquated term used in law and law enforcement, including the forgery and documentary fraud provision under s. 374 of the Criminal Code, R.S., 1985, c. C-46.

[40] Over the objection of the plaintiff, I admitted into evidence the Policy & Procedures of the BCPIO tendered by the defendant. This monograph contained the following description of the SIFT alert system:

This database is managed by the CBA and is in accordance with the submission to the Industry Canada, to be limited to employees who have been dismissed for fraudulent or other criminal activity. This would include employees terminated for serious breaches of Bank Code of Conduct but not for terminations for cause such as Incompetence.

Based on the admission of Conroy that there was no prohibition against accepting referrals from dual licensees (realtors and mortgage brokers) provided the involvement was limited to assistance to a purchaser in a transaction, the quality and seriousness of the allegation of procedural breach by Curtis in accepting referrals from one or more such licensees was at best subject to further inquiry before an escalation to SIFT. I will discuss this issue in greater detail when dealing with the qualified privilege defence. For now, based on the paucity of evidence tendered by the defendant to show that there had been fraudulent documents, I interpret the word “uttered” in the context of the gloss to the contents of the Roberts report as an attempt by the investigator to shoehorn Gary Curtis for inclusion in a database for known criminals. Had the request form cited failure to verify information or other forms of “Incompetence,” the request for the alert would not have qualified.

[41] As a result of the SIFT alert, Curtis’ name popped up when his prospective employers performed their background checks. The bank admitted that on separate occasions during May 2012, it sent the following response to CIBC on May 14, to RBC on May 16, and to BMO on May 24:

Advised that subject had submitted fraudulent documentation in support of mortgage applications.

[42] The bank conceded that the above statement was defamatory. Moreover, given the nature of the database, there could be no doubt that members submitting inquiries and receiving such a response would understand that the word “knowingly” could be inferred.

[43] On May 24, 2012, the investigator submitted her final report. The only outcome of the April 25 interview apart from a note that it was done was a two-line conclusion and recommendation: “We recommend that the business line sever relations with all parties involved in a manner prescribed by legal counsel. A list of parties identified will be provided by S&I under separate cover.” On May 30, 2012, a different senior S&I investigator reported various names from the Curtis investigation for inclusion in the bank’s fraud database.

[44] The bank did not report Curtis to the police and did not investigate his other files.

[45] During May 2012, RBC and BMO had extended provisional offers of employment which they withdrew after the SIFT alert and correspondence with the defendant. CIBC declined the employment application after a similar process. In effect, the plaintiff was blocked from employment of any kind in a mainstream financial institution.

[46] In March 2015, the plaintiff inquired with the BCPIO to find out the inquiry history. He also inquired again in 2019 and found out that the alert would be lifted after a 7-year period, but the defendant had always been able to remove the alert at any time.

[47] Before I leave this issue, I will comment on another ground cited by the plaintiff in his defamation case, the bank’s internal coding of his personnel file.

[48] On May 4, 2012, a payroll clerk asked for verification Curtis’ employment status before processing a request. Roberts responded to the request by confirming her authorization that he be coded “not eligible for rehire.” There was no evidence as to what the request had been. The plaintiff could have obtained this information in the extensive discovery process. This was the only information that could be construed as a possible communication outside the bank of his coded status. The plaintiff contended that this is information that other bank employers request in their background checks of prospective employees. The bank contended that these were internal inquiries to ensure they were not rehiring an ineligible candidate. The plaintiff’s evidence on this point was speculative. Truth is also a defence to the communication of this information outside the bank, even though the plaintiff has a point that it would be detrimental to his employment prospects.

[49] I find that there was insufficient evidence that the rehire ineligibility was communicated outside the bank to warrant expanding the defamation beyond the conceded communication to the BCPIO and to Curtis’ prospective bank employers. Moreover, I find that the evidence of the withdrawn employment offers in 2012 signified that the plaintiff’s prospects of gaining re-employment in the mainstream banking industry was essentially rendered barren for seven years.

[50] Had the SIFT alert not been requested unjustifiably by the bank investigator, the inference can be drawn, as invited by the bank in defence of the damages case, that the plaintiff could have accepted any of the three employment offers he received in May 2012, and that CIBC could have offered him a job as well.

[51] While the bank did not defame the plaintiff by recording internally that he was ineligible for rehire, the report to BCPIO for inclusion on the SIFT database, as well as the communications to inquiring employers that Curtis had submitted fraudulent mortgage documents, established prima facie liability on the part of the bank. The bank conceded this, and it relied on the two defences.

ISSUE #2 – JUSTIFICATION

[52] The defence of justification is simply a legal phrase for the common understanding that truth is a full answer to a libel suit. Technically, truth counters the presumption of falsity and eliminates the prima facie cause of action based on that presumption. Justification or truth is a positive defence which the bank had the obligation to prove on a balance of probabilities. It does not mean the defamer was right to express a falsehood. That is the domain of the defence of privilege.

[53] The justification defence can pose tactical problems because the utter inability to prove the truth of the statement can feed the malice exception to qualified privilege. The idea of “malice,” as commented by Cecil A Wright and Joseph Sedgwick in their review of Gatley on Libel and Slander, Third Ed., 1938 16-10 Canadian Bar Review 819, at 820, describes “an act done without legal justification.” Although the malice exception to qualified privilege requires actual malice, that comment reflects the pragmatic understanding that one who utters a falsehood in the absence of reasonable grounds can be found to have done so with actual malice, as opposed to the malice presumed from the defendant’s concession that the plaintiff had a prima facie libel case.

[54] Going into the summary trial, the defendant relied principally on an affidavit from Conroy, who was the bank’s Vice-President of National Mortgage Sales n 2012. He relied on the report from Roberts, based on the investigator’s interview, for the conclusion that Curtis had breached various bank policies and procedures including “creating false documents and knowingly submitting false employment documentation.” Insofar as the bank tailored its defence of justification to the libel claim, the evidence from its main witness was a conclusion based on Roberts’ impressions of the investigation interview she had witnessed and any post-interview discussions she may have had with the investigator. In other words, the bank’s summary trial evidence to prove fraud was going to be mixed hearsay and double hearsay.

[55] The defendant did not call the actual S&I investigator to give evidence because she was no longer working there and may have left the country. Moreover, it did not call anyone else from that unit or its current iteration to interpret the investigator’s notes or reports to explain the investigator’s notes which I allowed as business records. Given the onus to prove the truth of the statement that Curtis had knowingly submitted fraudulent documents, i.e., that he actively perpetrated fraud or was wilfully blind to it in his work, I did not understand how the defendant intended to prove justification. In fact, in cross-examination by the plaintiff yielded an admission from Conroy that he had seen no documentation of fraud and that he based his opinion regarding Curtis’ misconduct entirely on his trust in the fraud investigator and on Robert’s competence as an employee-relations professional.

[56] Going into the trial, the bank’s only witness had neither the means nor the qualifications to prove through admissible evidence that the libel was justified. Because of the summary trial format, I had the benefit of reading Conroy’s affidavit and knew the bank’s evidence deficit before the bank’s turn to present its defence. It was the plaintiff who summoned Roberts to testify and presented the defendant with the opening at least to fill in some of the gap in the defence. During several debates between the plaintiff and me over the plaintiff’s obsession with trial time allocation, the defence clearly gleaned from my attempts to discourage calling the bank’s employee under rule 53.07 – a risky and difficulty procedure even for experienced counsel – that the bank’s evidence on the justification defence was rather thin. Thus, when the plaintiff wisely heeded the court’s advice not to call Roberts, the bank’s lawyer expressed her intention to call that witness. The plaintiff then told me I cannot allow that to happen, and I overruled his objection.

[57] Conroy’s admission of an absence of documentary proof of fraud stands in contrast with the affidavit, serving as the defendant’s evidence in chief, attaching Roberts’ report concluding that Curtis had “breached the Guidelines for Business Conduct with respect to creating false documents by knowingly submitting false documentation.” The gap between the bank employees’ stated belief in what Curtis had perpetrated and evidence in support of the belief signified only that the bank had breached its own criteria for SIFT database handling. Those criteria required admissions of fraud or direct or strong circumstantial evidence.

[58] A case in point illustrates the difficulty in resolving the issue in the bank’s favour. One of the suspect mortgage applications entailed an application the plaintiff had allowed to be processed and approved by the underwriting team. The employer confirmation letter was stamped as approved, after the underwriter telephoned the business and verified the employee’s service and income details. During the S&I investigation, the fraud investigator telephoned the same number for the employer several months later and received a confused answer.

[59] As I ruled during the trial, the tables containing the investigator’s notes could be admitted as business record evidence that the investigator had made the call and obtained the recorded answers. The notes also recorded that she had visited the address of the business under scrutiny and found that it was only a mailbox at a UPS Store. However, I also declined to admit the note as proof of any non-transactional opinion or conclusion arising from the review of the documents and making of the call.

[60] In Adderley v. Bremner, 1967 CanLII 308 (ON SC), 1967 CarswellOnt 217, [1967] O.J. No. 1153, [1968] 1 O.R. 621, 67 D.L.R. (2d) 274, at paras. 8-9 Carswell, this court admitted hospital records for the purpose of routine medical procedures such as date and time of admittance, names of attending physicians, administration of drugs and nurses’ notes. The court refused to admit the records for the purpose of proving statements of opinion, diagnosis, or impression. More recently, the court came to the same conclusion in Baines v Abounaja, 2023 ONSC 2078, at para. 171. The standard of proof for the justification defence, i.e., whether the statement published by the defendant was truthful, is the civil standard of balance of probabilities. For a long time, proof of fraud commanded an elevated level of rigour: Gerling Global General Insurance Co. v. Siskind, Cromarty, Ivey & Dowler, 2004 CanLII 4856 (ON SC), at para. 219. This view changed after the Supreme Court’s statement in F.H. v. McDougall, 2008 SCC 53, [2008] 3 SCR 41, at paras. 45-46, that there can only be one rule of clear, convincing, and cogent evidence to satisfy the balance of probabilities test.

[61] As a practical matter, the burden of a defendant in a libel suit is hard to distinguish from that of a plaintiff in a civil fraud case. To justify the libellous statement, the bank must prove, on a balance of probabilities, that Curtis submitted for processing mortgage applications he knew contained false information for the purpose of deceiving the bank’s other employees into approving the said applications. The only substantive difference would have been that if the bank had been a plaintiff in a fraud suit it would also have had to prove harm suffered by the fraud, such as the costs of defaulted mortgage loans. It would be contrary to the purpose of fraud prevention and deterrence for the bank to prove the full cause of action for civil fraud, viz., proof of damages, to justify a SWIFT alert for a former employee caught submitting fraudulent documents. The bank’s burden consisted of two elements it had to demonstrate to justify the libel: (1) false information in the documents, and (2) Curtis’ submission of the documents for processing knowing of the false information.

[62] Roberts testified that the investigator’s call made months after the underwriter’s verification call during the original application process can uncover fraud, because the person on the other end of the line would not have been expecting the call. In contrast, a call at the time of the application could elicit a rehearsed or scripted response if the shelter fraud had been staged. Roberts thus presented evidence exculpating Curtis from the allegation that he had knowingly participated in the fraud, by demonstrating how the fraud could only have been uncovered a long time after Curtis processed the mortgage applications. This followed a repeated tug of war between the plaintiff and the court, whereby he constantly objected to procedures and admission of evidence that turned out to benefit his case and sought to embark on procedures and tender evidence in ways that made his case appear weak and contrived.

[63] Crime detection and investigation entails a degree of belief-based bias, a.k.a. reasonable grounds, or Bayesian probability: See the public bench book, Science Manual for Canadian Judges (Ottawa: NJI, 2018), at pp. 72-73. At first glance, the bank certainly could not have been faulted for investigating Curtis’ files. The investigator’s Initial Report stated that, a separate investigation led to a review of Curtis’ portfolio which “identified 12 mortgages which were established based on falsified supporting documents.” The matter was not reported to the police. Later in the report, the report stated that the 12 files were detected in a review of “15 randomly selected files.” The Final Report increased the number of reviewed sample mortgages to 18, with 16 containing fraudulent documentation. Had the investigator testified, the plaintiff could have cross-examined her on her methods, starting with the obvious difficulty of finding 4 additional suspect files after increasing the sample pool by 3. Perhaps there was an explanation for that.

[64] The bank needed to investigate Curtis to make sure it was not employing a fraudster. Neither the investigator’s report nor Roberts emerged from the interview with more than an opinion of his credibility. This served to confirm their belief, but they failed to realize there was still no evidence of his complicity. Given the seriousness of what they were about to do to him, confirmation bias was simply not good enough.

[65] Roberts testified that she concluded that Curtis must have been involved in the fraud after he gave inconsistent answers to questions about certain mortgage brokers, including one identified as “Mr. O.” and declined to correct them when given the opportunity to do so. After the interview, it turned out that Mr. O. was on a “do not use” list of mortgage brokers.

[66] Perhaps the bank could have proven fraud on a balance of probabilities by calling a witness with expertise to explain the investigator’s notes and methods. On the face of the investigator’s reports and Roberts’ report and testimony, it had not adduced sufficient evidence to prove that there had even been a fraud committed by the mortgage applicants or by their brokers. Increasing the probability of fraud is not proof of fraud but mere suspicion if the evidence does not enable the court to weigh the likelihood of fraud against the likelihood of explanations for the data short of fraud.

[67] Perhaps criminal organizations use private mailboxes and have bad telephone manner, but so do legitimate businesses. The lack of bricks-and-mortar premises and professional telephone manner are, from anyone’s perspective, grounds to doubt a business’ legitimacy. However, that doubt does not amount to proof of fraud. In this instance, the investigator’s notes, including but not limited to the example I cited above, cannot be used to draw an inference that there were, in fact, fraudulent documents. Judicial inferences are not to fill in holes, but rather to “connect the dots” based on a solid evidentiary foundation: Waldner v Visentin, 2019 ABQB 855, at para 56. Even if there was a valid reason for having failed to make the investigator’s testimony available to the court, the defendant bore the onus of convincing that there were fraudulent documents, as opposed to raising even a strong suspicion.

[68] The bank submitted that hearsay contained in the investigator’s notes and reports was admissible based on the principled approach to admission of hearsay. Hearsay is presumptively inadmissible unless it falls under an exception. The exclusionary rule can also be challenged if the evidence is sufficiently supported by indicia of “necessity and reliability”: R. v. Mapara, 2005 SCC 23, [2005] 1 SCR 358, at para. 15.

[69] I heard evidence that the investigator was now living in another country. I heard nothing of how the bank knew that, or why that information might have led to an inquiry of which country. In short, the defence provided no evidence of the measures taken to locate the former employee. Given that the court frequently hears evidence by video link – as it did with Ms. Roberts – the foreign residence of the witness is not an issue. There was no evidence of the investigator’s reluctance to testify, or of the inability to obtain commission evidence. Nor was there evidence from someone who had worked with the investigator and could interpret her notes and help the court understand the methodology. The written information of the investigator, shielded from cross-examination even through a fellow or current fraud investigator, cannot satisfy the first test of necessity for admission of the hearsay.

[70] As with any litigant, there is a duty to preserve evidence if one intends to rely on it at trial. Individuals and self-represented parties bear that burden, so there is no reason to excuse a national bank simply because it is an institutional litigant. In fact, this last fact should be reason why it might be expected to have procedures for preserving evidence. The logical way to do this could have been to propose the investigator to be examined for discovery, or to employ rule 36 to take the investigator’s evidence before trial, when she provided notice to the bank of her intention to leave. I appreciate that banks, insurance companies and other large entities are not necessarily organized in this way. But, to paraphrase Curtis, the banks have policies for everything, so why not this one?

[71] Even if I am wrong about the necessity point, I do not see how the reliance on the investigator’s opinion by Conway or Roberts, neither of them employed in the S&I department, could have any probative value. The bank’s submission was that the court can rely on the hearsay of the investigator because the two other employees did. The principled approach does not allow the court to trust a witness’ trust in what another person says, where the other person is not before the court. Rather, the principled approach is satisfied in rare cases where the evidence, not the person who stated it, is trustworthy or capable of being tested. As stated by Karakatsanis J. in R. v. Bradshaw, 2017 SCC 35, [2017] 1 SCR 865, at para. 27:

The hearsay dangers can be overcome, and threshold reliability can be established by showing that (1) there are adequate substitutes for testing truth and accuracy (procedural reliability) or (2) there are sufficient circumstantial or evidentiary guarantees that the statement is inherently trustworthy (substantive reliability).

[72] For the bank to justify the libel, it needed to have cogent and convincing evidence that Curtis was involved in the fraud as a perpetrator and not as a victim of it as a bank employee. The business records introduced showed that, if the investigator had been called to testify, the bank could likely prove that the mortgage files contained fraudulent documents. In her absence, another investigator could have provided evidence of threshold reliability of coming to that conclusion. The bank’s evidence did not show that the bank had the basis to leap from that proof capability to an ability to demonstrate Curtis’ complicity. It is not for the court to speculate how the bank could have obtained that capability, but one would have been to call the police. Perhaps the police would have had the means to tread further into Curtis’ zone of privacy. The bank chose not to report the alleged fraud to the authorities. My conclusion on the point of principled hearsay is that the defendant did not do nearly enough to probe into the question of Curtis’ involvement to convert the hearsay of the investigator’s belief or opinion into a reliable source of facts.

[73] The further problem with the defence of justification is that the investigator herself did not actually reach the conclusion that Curtis had knowingly submitted the fraudulent documents. Her final report of May 24, 2012, identified the issue that Curtis “may be knowingly submitting mortgage applications referred by other mortgage brokers that contain falsified information.” The closest she came to confirming that hypothesis was that “S&I identified a mortgage processed by Curtis based on known fraudulent employment information.” Apart from the identification of a single file with such information out of 15 (not the 18 on the previous page), she did not specify whether Curtis knew or whether only “S&I” knew. That investigation took place on April 24, the day prior to the interview, thus suggesting the latter. Under the heading, “FINAL REPORT,” the investigator stated: “The matter will not be reported to the authorities.” Later, however, in preparing the SIFT alert requisition, she stated that “Strong evidence was found that Curtis knowingly uttered fraudulent documents on several applications.”

[74] The investigator’s input request to BCPIO simply copied and pasted Roberts’ report and did not show the investigator had come to an independent conclusion that Curtis had knowingly submitted fraudulent application documentation. Perhaps there could be an inference that the investigator agreed with that conclusion. However, such an inference does not rise to actual proof of Curtis’ prior knowledge of the fraud. Rather, the argument circles back to the court’s ability to rely on the investigator’s possible inference by way of trust. The hearsay rule does not allow this court to rely on such trust.

[75] In fact, the copy-and-paste nature of the investigator’s submission to BCPIO signified a procedural laziness that undermined the integrity of her conclusion. In the judicial context, this type of practice calls into question whether a trial judge came to grips with the issues and dealt with them independently and impartially: Cojocaru v. British Columbia Women’s Hospital and Health Centre, 2013 SCC 30, [2013] 2 SCR 357, at para. 18. The investigator was under no judicial duty to be independent or impartial, but if she and Roberts were bouncing reliance off each other’s conclusions, their reports cannot be characterized as reliable for the purposes of the principled admissibility of hearsay. In a civil trial, the fact that the bank produced no report authored by the investigator reasoning and concluding that Curtis knowingly took part in the fraudulent mortgage applications meant the bank had no basis to defame him.

[76] Roberts’ explanation of the investigator’s ability to uncover fraud by following up several months after the approval of the application also indicated that Curtis had been in no better position than the underwriter to detect the fraud. Indeed, if Curtis relied on the underwriter to verify the employment status of the mortgage applicant, it would have been even easier to pull the wool over Curtis’ eyes. Instead of proving Curtis’ knowledge of the fraud, the bank’s evidence proved he likely did not know and did not have the means of knowing about the fraud. Roberts’ evidence extends more generally to the entire interview of Curtis, in that she appeared to have paid close attention to his allegedly evasive responses to findings presented by an experienced fraud investigator and his explanations why he failed at the time to verify the legitimacy of the information from the brokers or the customers.

[77] What is left in the bank’s evidence of justification was the report and oral evidence of Roberts, who sat in on the interview with the fraud investigator from S&I and took notes. Much of the interview as reported by Roberts dealt with the plaintiff’s dealings with mortgage brokers. They focused particularly on his dealings with one mortgage broker one “Mr. O.” Based on Curtis’ confusion or discrepancies in his statements over the length of time he had known this broker, Roberts decided his responses to questions “were not credible and it is our conclusion that Mr. Curtis knowingly submitted fraudulent documentation for Mr. O’s personal refinance mortgage.”

[78] Of Curtis’ dealings with another broker, she wrote that he “accepted and submitted fraudulent and forged documentation and knowingly breached the Bank’s Policies and Procedures with respect to obtaining documentation.” Given the seriousness of the matter, she knew or ought to have known that this did not amount to knowingly accepting and submitting the fraudulent documentation. Indeed, Conroy’s evidence about dealings with dual licensees left some doubt whether Curtis had even breached bank policies and procedures regarding referrals from mortgage brokers.

[79] The plaintiff repeatedly pointed out that Roberts, whose role was to act as a human resources bridge between the employee and the bank’s investigating unit and mortgage department, had no expertise in fraud investigation. She had been involved in about 50 of these investigations of employees. Although she should not be faulted for lacking detailed expertise on the mortgage application process, her conclusion that the employee had knowingly submitted fraudulent mortgage application documents based on his lack of credibility on one file and on her belief in his wider disregard for bank procedures was at best a logical leap. Given the grilling Curtis received over the referrals from mortgage brokers against bank policy, as well as the amount of evidence the bank tendered on the subject during the trial, an impartial observer of Curtis’ interview by the S&I investigator would have thought the bank was going to reprimand him for being careless in accepting applications through the mortgage broker channel.

[80] The bank’s witnesses both stressed the importance of refusing being courted by mortgage brokers. There was a separate unit at the department that dealt with them. Mortgage brokers could only obtain a commission if they dealt with that unit, rendering suspect any mortgage broker that would court an MDM. The plaintiff’s response to this was that these were dual licensees who were acting on house sales as realtors.

[81] The plaintiff testified about a multi-stage process for mortgage application approval. His job description did not include verification of income and employer information. Conroy testified that everyone in the process, including MDM’s were responsible for the integrity of the process. The plaintiff demonstrated, through discovery read ins and from cross-examination of the defence witnesses, that all of the approved files had undergone verification by other employees, including those with the specific mandate to check up on the accuracy of the information. Roberts admitted in her discovery read ins that the underwriting unit was “the one that’s verifying the income and down payment” information in every one of the files.

[82] Roberts’ evidence of how the fraud investigator could trip up a fake employer by telephoning months after an application demonstrated that an MDM such as the plaintiff, being the first point of entry into the mortgage unit, was probably the employee who could mostly easily be duped by a shelter fraudster. There was no evidence that anyone else down the mortgage assembly line was ever accused of fraud for having approved one of the plaintiff’s customer’s applications.

[83] I found it unsettling that a major Canadian bank would rely on a commissioned sales employee to be its principal bulwark against mortgage fraud and similar risks. As the plaintiff established when he cross-examined the bank’s witnesses, nowhere in the MDM job description did the bank include income and employment verification as a specific area of responsibility. The defence attempted to smooth over this by referring to his relationship-building tasks, but all of these concerned the development of customer networks with a view to growing the bank’s mortgage business.

[84] When the plaintiff cross-examined Conroy about the role of a mortgage underwriter in performing income and employment verification, Conroy attempted to discount that role by characterizing such members of the mortgage staff as junior. This did not help the bank’s case. If one believed Conroy’s evidence, the employee incentivized to bring in as many mortgages leads as possible was the one tasked to evaluate and investigate risk of fraud, and the salaried employee whose job description was to evaluate and investigate risk of fraud was not sufficiently experienced or qualified to perform that task.

[85] As a matter of business process, investing more in sales than in underwriting is an institutional priority. I accept that a bank’s mortgage department cannot employ fraud investigators to scrutinize every mortgage application that comes in. Underwriters also do not bring in new banking clients. If the bank truly expected Curtis to detect fraud better than underwriters, this would have been a subject of training and/or discipline. It was not a justification for putting him out to be pilloried in the market square as a banking fraudster.

[86] My appreciation of the evidence at trial was that the interrogation by the fraud investigator likely frightened Curtis that he faced reprimand or dismissal over his reliance on a network of dual real estate and mortgage broker licensees. The investigator may also have expressed concern about the financial risks to the bank resulting from potential delinquencies or defaults by over-leveraged mortgagors. The heavy emphasis placed by the defendant on the issue of Curtis’ flouting of risk management guidelines only strengthened the overall impression that the bank suspended him for failing to protect the bank from financial risk, and not for being the actual source of the risk.

[87] The court’s rejection of the defence of justification for failure to prove that Curtis knew of the fraudulent information in the mortgage applications should not, in this instance, leave the defendant wondering about the level of proof required to justify their defamation of their former employee. This question nevertheless illustrates the issue overlap between justification and qualified privilege, in that the bank’s SIFT manual provided criteria for entry into the database:

1. Admission of criminal fraud

2. No admission but agreement by the employee to pay for the loss

3. No admission but direct evidence of fraud

4. No admission but strong circumstantial evidence;

5. Termination for cause.

[88] The last point is interesting because Roberts recommended termination for cause, had Curtis not resigned. Had he not resigned, I doubt that he would have been terminated for cause without a referral of the question to legal counsel. The bank’s criteria reflect the high self-imposed bar for proof of an offence warranting entry to the database. The bank elected not to refer him to police, even though law enforcement would have been the obvious choice to escalate the intrusion level necessary to obtain admissions or to conduct a sting operation. The only motive given by Roberts for the fraud was about $22,000 in commissions earned on the suspect mortgage files out of total income of $285,000, although these data could just as easily have been evidence that Curtis did not have much of an incentive to jeopardize his lucrative career for an 8% margin.

[89] The best answer to the question of the evidence required to justify the defamation is that fraud is necessarily hard to prove because of the intent to deceive the victim. The bank, as an institutional victim, must shoulder the responsibility to devote sufficient investigative resources. The weakness of the evidence presented at trial, as well as the absence of efforts to preserve the proof backing up the defamation, was on display as the bank attempted to prove its defence by attempting to adduce indirect and unreliable hearsay.

[90] Based on the evidence at trial, I could not make a finding that what the investigator found to have been fraud in various files in fact amounted to fraud. The evidence of Conroy and Roberts that they trusted the investigator to perform that fact-finding task was reasonable, given their roles in the organization. However, in telling the court that it should trust the investigative findings because Conroy and Roberts trusted them, the institutional defendant knew very well that this second-hand evidence did not make the cut for the purposes of the justification defence.

[91] The defendant therefore pleaded and raised justification and thereby bore the onus of proving fraud, and of proving the plaintiff knowingly submitted the fraudulent documents for the mortgage underwriters’ approval. This defence failed because there was insufficient evidence that there was fraud and no evidence that the plaintiff was ever aware of it until the files were presented to him at the interview. Indeed, the evidence of the investigative methods employed to detect fraud proved, on the balance of probabilities, that neither the plaintiff nor the underwriters charged with the duty to verify the information would have been positioned to detect the alleged fraud. The justification defence therefore fails.

ISSUE #3 – QUALIFIED PRIVILEGE

[92] The BCPIO is a division of the CBA, a voluntary organization of the major Canadian chartered banks and qualified financial services entities. The defendant relied on the role of the BCPIO in protecting the public and the financial services industry from white collar crime. Its mission and mandate were stated in the User Guide as follows (italics mine):

The mission of BCPIO is effective information sharing for the prevention, detection, investigation, and deterrence of financial crime against its member institutions, their employees and the clients they serve.

The mandate of BCPIO is to enable designated employees of BCPIO member organizations to collect, use and disclose personal information in the prevention and investigation of criminal and dishonest activity that contravene laws of Canada, a province, or a foreign jurisdiction. The BCPIO provides a controlled and secure environment in which information about a fraud or crime can be quickly and safely exchanged between BCPIO members to maintain the security and integrity of our banking system. The collection, use and disclosure of personal information is done in accordance with PIPEDA.

[93] Once the alert prompted the plaintiff’s prospective employers to inquire with the defendant, the defendant informed them that Curtis had submitted fraudulent documentation. The defendant conceded that the combination of the alert and the follow-up advice was defamatory. The defendant asserted the defence of qualified privilege on the basis that the bank owed a public duty to report fraud to the BCPIO and to express an honestly held belief in a reference to a prospective employer.

[94] Although stated in the context of an anti-SLAPP motion under s. 137.1 of the Courts of Justice Act, R.S.O. 1990, c. C.43 (CJA) to stop defamation suits thwarting open debate, the following description of the defence of qualified privilege by Côté J. in Bent v. Platnick, 2020 SCC 23, [2020] 2 SCR 645, at paras. 121-22, reflects the settled law on the subject:

121. An occasion of qualified privilege exists if a person making a communication has “an interest or duty, legal, social, moral or personal, to publish the information in issue to the person to whom it is published” and the recipient has “a corresponding interest or duty to receive it”: Downard, at §9.6 (footnote omitted). Importantly, “[q]ualified privilege attaches to the occasion upon which the communication is made, and not to the communication itself”: Hill, at para. 143; Botiuk, at para. 78. Where the occasion is shown to be privileged, “the defendant is free to publish, with impunity, remarks which may be defamatory and untrue about the plaintiff”: Hill, at para. 144; Botiuk, at para. 79. However, the privilege is qualified in the sense that it can be defeated. This can occur particularly in two situations: where the dominant motive behind the words was malice, such as where the speaker was reckless as to the truth of the words spoken; or where the scope of the occasion of privilege was exceeded (Downard, at §1.9; see also Hill, at paras. 145-47; Botiuk, at paras. 79-80).

122. For this reason, a precise characterization of the “occasion” is essential, as it becomes impressed with the limited, qualified privilege, which in turn becomes the benchmark against which to measure whether the occasion was exceeded or abused.

[95] The plaintiff did not concede that the BCPIO and the employer reference communications could have given rise to qualified privilege. The Court of Appeal has stated that they are occasions for the operation of the privilege, in the following description of the common-law rationale for it:

[39] Employment references, business and credit reports, and complaints to police, regulatory bodies or public authorities are classic examples of occasions of qualified privilege. The rationale for qualified privilege is that on such occasions, "no matter how harsh, hasty, untrue, or libellous the publication . . . the amount of public inconvenience from the restriction of freedom of speech or writing would far outbalance that arising from the infliction of private injury" (Huntley v. Ward (1859), 6 C.B. (N.S.) 514, at p. 517).

[40] The privilege is said to be qualified as it can be defeated upon proof of malice, that is spite or ill-will, ulterior purpose, or, more commonly, proof that the defendant either knew the statement was false or was reckless as to its falsity: see Hill v. Church of Scientology of Toronto, 1995 CanLII 59 (SCC), [1995] 2 S.C.R. 1130, [1995] S.C.J. No. 64, at para. 145 ("Hill v. Scientology"). The privilege can also be lost where the limits of the duty or interest are exceeded by the use of words not reasonably appropriate to the occasion.

Cusson v. Quan, 2007 ONCA 771, at paras. 39-40

[96] All of defamation law involves such apparently contradictory rationale, because of the use of shifting onus that ultimately and clearly favours protection of reputation over freedom of expression: Cusson, para. 37. The defence urged that para. 39 above provided the bank an almost blanket protection. I do not read the paragraph in that way. The Court of Appeal intended to explain the rationale for the breadth of the privilege as an exception to the ordinary outweighing of reputation over expression. When read in conjunction with para. 40, I do not read para. 39 as a licence to defame, where the evidence supporting the publication fall well below the criteria for exercising the duty to report. Moreover, the publication has a lasting effect of being republished automatically every time a prospective employer makes an inquiry. In that instance, there is a continued duty to ensure the ongoing publication is not driven by malice and that the publication meets the criteria for exercising the duty.

[97] The qualified privilege may be defeated before considering the malice exception if the defamation exceeded the limits of the duty or interest occasioning the privilege. An important feature of the occasion is the necessity to communicate the statement to the party with the corresponding duty to receive it: Bent v. Platnick, 2020 SCC 23, [2020] 2 SCR 645, at paras. 128-30. The occasion is defined by the existence of the duty in the specific instance, and not a duty at large: Botiuk, at paras. 82-85. This idea is key: if the duty is not engaged, there is no privilege.

[98] The BCPIO was a possible forum or “occasion” for banks to share information about financial crimes and those who commit it. The unit has developed an alert system called SIFT. Member banks can submit the names and basic biographical information of individuals. If a member receives an application for employment from an individual whose personal information matches one on the list, it triggers an alert. The potential employer can then request the nature of the alert submission from the member that requested the alert. Inclusion in the register lasts for as long as seven years, unless withdrawn by the listing member.

[99] As I mentioned earlier in the analysis of the justification defence, the criteria for inclusion require either an admission of criminal or quasi-criminal wrongdoing, direct evidence, strong circumstantial evidence, or termination for cause. The principles of qualified privilege and justification overlap in that the protection that qualified privilege affords the defamer is only as strong as the credibility of the belief in the truth of the statement. Whereas absolute privilege is not linked to the truth or necessity of the utterance, such as statements made in a legislative or judicial setting, qualified privilege requires the grounds for belief in the statement to be connected to the forum’s criteria for making the statement.

[100] The fact that the bank’s S&I unit summarily decided not to report Curtis to the police was evidence it considered SIFT to be sufficient to protect CBA members. It was also evidence that the bank did not want the police to investigate its mortgage files. Biblical scholars will recognize the practice of casting a community’s sins on the head of a goat and releasing it to starve in the wilderness.

[101] The purpose of the BCPIO is not to serve as a bulletin board for banks to ban employees from employment based on incompetence or other occupational deficiencies. It has been described as a database for reporting “individuals found guilty of serious banking crimes”: Joshi v National Bank of Canada, 2016 ONSC 3510, at para. 4. It is not limited to former employees but to others who are involved in committing financial services fraud: Norama Design Inc. v. CGU Insurance Company of Canada, 2007 CanLII 17034 (ON SC), at para. 17-18. These findings are not binding on me or on the defendant. However, the interpretations of the BCPIO mandate are consistent with its use to flag criminal operators. As an alert, only basic personal information goes out to inquirers. Because of the seriousness of the wrongful activity required to put an individual on the register, the rationale for reporting is therefore one inflection point for the libel and qualified privilege defence.

[102] The BCPIO mandate in relation to financial crime therefore defines the occasion, for the purposes of qualified privilege. If the bank had evidence of fraud by persons other than Curtis, it would have had reason to report those persons to SIFT. There was no evidence that it did report the actual fraudsters to SIFT. Rather, the investigator recommended that “business lines sever relations with all parties,” understood to mean the cancellation of mortgages, and provided a “list of parties identified.” The bank concluded that its investigation of Curtis was an occasion for reporting him to SIFT without grounds to do so according to BCPIO criteria. It did not report the known perpetrators to the same database, even though page 5 of the BCPIO User Guide referred to the exchange of information “on a client or individual involved, or suspected of being involved, in criminal activity.”

[103] Based on the investigator’s reports and Roberts’ report, they knew or ought to have known that nothing uncovered in the investigation could have informed the use of the words “Curtis knowingly uttered fraudulent documents” in the SIFT Database Input Request Form. Put another way, if they had indeed uncovered evidence that Curtis had uttered such documents, they certainly did not note it in their reports. The use of inflammatory and unfounded characterizations to justify inclusion in the SIFT database must count as the use of words not reasonably appropriate to the occasion of the qualified privilege. Even if the CBA operates the BCPIO on an honour system without verification, a justification that Curtis had breached bank policy by taking referrals from mortgage brokers would not have qualified him for inclusion in the database and would have at least provided grounds for the plaintiff to request expungement when he later contacted the BCPIO.

[104] Indeed, the BCPIO Policies and Procedures handbook also contained a chart outlining the types of activities the BCPIO expected members to report, all of which were forms of fraud or defalcation. The category for lack of lending due diligence, including the failure to “validate the clients provided assets, liabilities, or income information in order to obtain approval on credit application - Could be single event or habitual behaviour,” the manual criteria stated, “NO SIFT,” meaning that the BCPIO did not expect members to report such individuals.

[105] The distinction between employee diligence or performance and the type of behaviour warranting inclusion in the SIFT register is obvious. A reprimand or even dismissal allow for an employee to dust himself off, reflect on what he has done or failed to do, and pursue retraining or some form of beneficial self-help. In the plaintiff’s instance, the fact that he was a top producer in the BNS mortgages department could counteract any disciplinary issue, such as the inappropriate handling of referrals from mortgage brokers. A prospective employer can weigh the risks and the benefits in considering him for hiring. Indeed, Conroy’s evidence at trial established that this unit of the bank valued sales more than underwriting in terms of hiring experienced personnel. A statement that Curtis had committed fraud, however, would have rendered him untouchable.

[106] An allegation of criminal behaviour, especially on a register intended to flag criminals, must be made only after the exercise of great restraint, including a duty to wait until an investigation is complete: Hill v. Church of Scientology of Toronto, 1995 CanLII 59 (SCC), [1995] 2 SCR 1130, at para. 155. This rationale is equally applicable to law and to other professions and callings: Botiuk v. Toronto Free Press Publications Ltd., 1995 CanLII 60 (SCC), [1995] 3 SCR 3, at para. 92. There is no principled reason for applying a conservative approach to qualified privilege in the instance of lawyers but not to bank employees. Both are callings that depend on reputation and integrity.

[107] The final report from the investigator was completed after the April 25, 2012, interview. However, it did not mention the interview and did not make any findings from it. As stated above, the investigator’s SIFT requisition was based on the human resources manager’s report and not on her own. What is most wanting in the piece was the absence in the reports of any real sense that the subject knew he was being investigated for knowingly having submitted fraudulent documents – the conclusion they reached and the basis of the fraud report to BCPIO.

[108] The defendant urged on me that the court’s role is not to rely on its own fact-finding and charge the defamer with recklessness in support of the plaintiff’s malice rebuttal to a defence of qualified privilege. I agree that to do so would collapse the justification and privilege defences. The failings of the investigation in the bank’s rationale for making the SIFT alert do not inform the malice exception to qualified privilege. Rather, the issue as described by the Supreme Court decisions on the subject relate to the occasion of the privilege: whether it was necessary for the defendant to alert the BCPIO of Curtis, and then to inform his prospective employers that he had committed fraud, based on Roberts’ finding that Curtis’ responses to the investigator’s questioning was not credible. Unless the investigator had put the specific issue of his knowing involvement to him, Roberts’ opinion about his credibility or about the ultimate issue would have been irrelevant and certainly unnecessary. At best, her impressions about his credibility could have been relevant to the investigation if Curtis had been asked about his knowledge. This idea is not limited to the judicial rule of Browne v. Dunn. It is a basic idea of fairness among human beings.

[109] Had the investigator, mandated to investigate, concluded that Curtis had knowingly engaged in the frauds based on evidence of such knowledge or even prevarication upon confronted with the accusation, the BCPIO criteria required her to assess whether the evidence met the criteria for reporting Curtis to SIFT. If not, she could still have investigated Curtis and asked him to come in again, during his notice period. If, at some point, she gathered enough evidence of his complicity to report him, the occasion would have presented itself. In such an instance, the qualified privilege protects the ability of a duty-bound person to communicate the information even if it turns out to be false. Thus, the fact that the court can reject the justification defence does not mean the qualified privilege defence must fail. It can still succeed, provided the facts establish the occasion for the privilege.

[110] If the evidence of the investigation showed at its highest that Curtis broke bank policies and failed to exercise due diligence, the duty to report him to SIFT did not arise. Perhaps if the investigator had testified, the bank could have proven that such a duty arose. However, the hearsay evidence of her investigation and Roberts’ interpretation of Curtis’ credibility fell short of establishing even grounds to believe that he had knowingly committed fraud turned a blind eye to it. The criteria in the manual even stated that there was no duty to report him to SIFT based on the actual and objective findings in the investigation. The defence of qualified privilege must fail because the duty to communicate the information did not arise.

ISSUE #4 – THE MALICE EXCEPTION TO QUALIFIED PRIVILEGE

[111] The rejection of the qualified privilege defence means the plaintiff is relieved of the burden to show the libel was malicious. In case I am wrong about the scope or application of the privilege, I will address the issue of malice.

[112] The plaintiff’s evidence regarding actual malice in reporting him to BCPIO and to alerting his potential employers was entirely speculative. There was evidence that he may have missed out on a perk in the first quarter of 2012 because of an error by his regional manager. There was evidence that the bank charged his personal accounts for penalties incurred by mortgage clients. None of these rose to the level of actual malice and were consistent with errors made by bank employees. There was no evidence that Curtis brought this error to the attention of his superiors at the time.

[113] Malice can defeat qualified privilege. It need not be actual malice. Actual knowledge of falsehood or reckless disregard for the truth can defeat privilege, but bona fides is presumed to exist: Botiuk, at para 79. The New Brunswick Supreme Court, Appellate Division in Culligan v. Graphic Ltd., 1916 CanLII 885 (NB CA), 1917 CarswellNB 1, 37 D.L.R. 134, 44 N.B.R. 481, at para. 8 Carswell:

If there is to be an investigation into the motives of the person publishing the libel, it is essential to know all the facts; and it is obvious that, if the information upon which he acted was procured from a person or persons who could not possibly know anything about the matters in question, and he nevertheless published the statements complained of as if they were based upon sufficient information, that might be cogent evidence of malice.

[114] Knowledge of falsehood or reckless disregard are not the same as communications exceeding the occasion of privilege. The distinction may seem blurred in some instances, and this one may be one of them. The following summary of the treatment of recklessness in the malice exception appears in Teskey v. Toronto Transit Commission, 2003 CanLII 23353 (ON SC), which I have paraphrased from paras. 14-19:

1. Defendants are reckless if they published the allegations without considering or caring whether they were true or not.

2. Similarly, the defendants were reckless if they failed to make enquiries, as it was reasonable for them to do in the circumstances before publishing the allegations.

3. Recklessness is distinct from carelessness, impulsiveness or irrationality in arriving at a positive belief that the allegations published were true. Carelessness is not evidence of malice.

4. The specific experience and training of each individual defendant should be considered. Recklessness is not to be determined by the objective standard of a reasonable person. Recklessness is subjective, yet at the same time contextual and fact specific.

5. The contextual factors to be considered include both the occasion giving rise to the Statements which include the facts and circumstances that existed at the time and the type of statement made.

[115] Counsel for the bank also urged on me that, unlike justification, the inquiry does not entail a re-investigation by the court. I agree. The recklessness element of malice requires an equivalent to knowledge of falsity, in the same way that Roberts equated in her report on Curtis’s knowledge of fraud and turning a blind eye to it. The bank’s investigator, before requesting the BCPIO to register him on the SWIFT database had to have known or she ought to have known that her grounds were false.

[116] When I say, “ought to have known,” I do not import an objective standard. It must be a subjective standard insofar as the investigator’s belief is the at issue. Subjective knowledge or belief is a difficult area of the law, but the courts have developed clear guidelines for dealing with it: R. v. Ewanchuk, 1999 CanLII 711 (SCC), [1999] 1 SCR 330, at para. 29.

[117] The problem posed by defence counsel’s correct characterization of the issue is that, as Ewanchuk illustrated, the state of mind of the subject is only provable through the subject’s testimony. Because the investigator was not called and her evidence was not preserved through processes available in the court’s rules, this presents an evidentiary problem. But for the legal presumption of bona fides, the validity of the investigator’s subjective knowledge or belief can only be introduced in evidence if the subject testifies. The presumption wears rather thin, if grounds for the honest belief are absent in the written report of the subject and the subject does not testify.

[118] The sum of the evidence from the bank, as I discussed in the justification defence, was that the bank had no cogent reason to know that Curtis had committed fraud. Therefore, the question must turn to whether the bank, acting through its employees, knew or subjectively ought to have known that there was no evidence of fraud by Curtis but communicated that falsehood despite that knowledge. The analysis of any proposition involving “false” answers in conjunction with negative propositions is fraught. The problem here, of course, is the absence of cogent evidence proving the commission of fraud. Roberts’ doubts about Curtis’ credibility did not rise to positive evidence of this serious offence.

[119] The exercise of drawing adverse inferences is discretionary and case specific. It should not be drawn unless warranted: Parris v. Laidley, 2012 ONCA 755, at para. 2. In civil cases, the failure to call evidence may amount to an implied admission that the evidence of the absent witness would be contrary to the party’s case or would not support it: R. v. Jolivet, 2000 SCC 29, [2000] 1 SCR 751, at para 28. On balance, I did not sense of a deliberate failure to call the investigator or make other arrangements to get the corporate evidence of its former employee’s fraud before the court. The court cannot draw an adverse inference to punish corporate indifference to barristers’ and courts’ obsession with the cataloguing of proof. That said, the plaintiff would be quick to remark on the unfairness of excusing the bank of the application of the rule of evidence when it had ample resources to be advised by an army of in-house lawyers and external counsel.

[120] The absence of an adverse inference against the defendant that the investigator knew or subjectively ought to have known there was insufficient evidence that Curtis was guilty of fraud, of “uttering,” means the absence of evidence of recklessness does not help the plaintiff, who has the substantive burden of the malice exception to qualified privilege. The belief of Roberts, who was not the fraud investigator, is of marginal relevance but favours the defence. The bank’s treatment of Curtis as a scapegoat by casting him out and reporting him to BCPIO while shielding itself from police or regulatory investigation by not reporting him or any of the files to the legal authorities could be evidence of actual malice. This could have been a subject area for the plaintiff to have pursued better at the discovery stage or in cross-examining the bank’s witnesses. However, his proclivity to aim for a knock-out blow without setting up witnesses with jabs made it easy for the witnesses to deflect the impact of his questions.

[121] Despite the weakness of the evidence of malice, partly due to the burdens on a self-represented litigant without courtroom training, there is a profound tactical unfairness in allowing the defendant, an institution with considerable legal resources, to mail in a defence and shield itself from scrutiny when it is the one relying on the qualified privilege defence. Despite this misgiving, I find as a fact that, at the time the investigator placed his request to the BCPIO for the SIFT alert, there was insufficient evidence that she and Roberts knew that Curtis had not committed fraud or were reckless as to the truth or falsehood of their conclusion. I also find that the likelihood that Curtis was used as a scapegoat for the underwriting failings of the bank’s mortgage unit was not sufficiently presented through evidence to amount to malice that would oust a qualified privilege defence.

[122] This does not end the malice inquiry, however. This case is unlike most libel cases, in that the publication is a passive communication that prompts inquiry. By its nature, the defamation pulls communication as opposed to casting it out. This distinguishes it from other publications, such as a traditional newspaper publication or radio-television broadcast. Given that the alert lasts for seven years unless withdrawn, it is more akin to a web publication that can be accessed through search engines for an indefinite length of time. After a libel notice or lawsuit, the publisher can investigate the truthfulness of the publication, to remove the webpage from public access, and to consent to withdrawal from search engines.

[123] Curtis’ resignation was a blunder, because it made it too easy for the bank to close the investigation before completion and to assume the worst of him. Running away always risks an inference of guilt. But then it is not every bank fraudster who sues. His lawsuit should at least have cancelled out the inference that he had tried to outrun the investigation.

[124] I must consider the bank’s lack of diligence in preserving the evidence of the investigator’s knowledge reflected a general failure take the plaintiff’s case seriously from the outset, including the admitted harm it caused. In the ordinary course of a defamation case, as typically launched against a media company, the publisher has an opportunity to retract a statement and mitigate the harm. Traditionally, that meant a libel published on the front page of a newspaper could be mitigated by a retraction and apology buried on the second or third page. Nevertheless, it meant something, and most libel lawyers would advise their clients against launching a suit.

[125] The court heard and read evidence of Curtis’ efforts to get the alert lifted. During the seven years that the alert acted as an impediment to Curtis’ employment in the mainstream banking sector, he was suing the bank, and the bank was defending the suit. During those years, the bank continued to defame the plaintiff and at no time did it lift the alert. Based on the evidence at trial, the alert was unjustified, and the Bank did not take measures to obtain the evidence from the investigator to determine whether it was or remained justified. This is therefore a unique defamation case in which the duration of the publication is not ephemeral as a news cycle or untargeted as in an old web story. Rather, the SIFT alert was a constant source of direct harm to the plaintiff. The failure to evaluate the paucity of evidence and take measures to retract the defamatory posting constituted an insensitive disregard for the harm it had caused. This was not malice after the fact but ongoing malice, at least at some point when it ought to have known that proving truth was beyond its reach.

[126] Knowledge of falsehood can also be a factor in the malice analysis after the fact, if there has been an unreasonable failure to issue a retraction: Wan v Lau, 2016 ONSC 127, at para. 47. I witnessed during the trial how difficult it must have been for the defendant to defend the litigation and litigations instituted by the plaintiff. Self-represented litigants, especially in the employment law context, tend to annoy employers and their lawyers. Nevertheless, as a regulated business founded by a public patent, the defendant has an obligation to uphold the law in all its iterations, including the law of defamation. At some point during the first seven years after the original defamation, the defendant had to have taken a hard look at the plaintiff’s muti-million-dollar lawsuit and considered whether they should take it seriously or consider it a nuisance. The concession of the prima facie case and the utter weakness of the justification defence meant the defendant could only cling on to privilege. That privilege started to erode once he sued, and the bank lacked a fresh investigation into the grounds for having defamed him.

[127] Taking the suit seriously included whether it had the evidence to defend the suit based on justification. At some point, the lack of the basis to prove even a subjective belief of the investigator had to solidify into actual knowledge that the defamatory statement was untrue. Not only did the bank fail to re-investigate the justification defence, but it also failed to perform a diligent search of its records in the litigation. The bank announced at the opening of trial that it had, after a dozen years of litigation, located the audiotape of the Curtis interview. This prompted the expected objection from the plaintiff.

[128] Whereas qualified privilege can protect a hasty or zealous belief if it turns out to have resulted in a false statement, the victim’s active protest and questioning of the basis gives rise to corporate knowledge that the libel was untrue or at least that its truth cannot be proven in court. In that instance, malice applies to the bank’s failure to retract the libel when it continues to act as a disparagement of the plaintiff’s employability. Although the failure to provide an apology is not, in itself, evidence of malice, an unfounded pleading of justification right up to verdict can be Halls v. Mitchell, 1926 CanLII 19 (ON CA), at paras 122-23. When the publication remains an active bar to employment prospects, the ongoing unjustified libel carries with it the accrual of malice to the point that it defeats the privilege.

[129] Therefore, even if I am wrong about the bank exceeding the limits of its qualified privilege, I find that in failing to remove the alert once the plaintiff brought his libel suit and in failing to preserve any evidence in support of justification, the defendant demonstrated disregard for the truth or falsity of the libel. Even if there was no malice at the time of the initial publications, this conduct constituted recklessness for the purpose of the malice exception to qualified privilege.

[130] Having concluded that neither of the affirmative defences have ousted the bank’s liability, I find the defendant liable to the plaintiff for having defamed him. In the law of libel, damages need not be proven, but the plaintiff must present convincing evidence of damages if he is pursuing damages beyond modest and discretionary amounts.

ISSUE #5 – DAMAGES

[131] The plaintiff’s suit consisted of the following claims, according to the latest amendment, re-ordered for ease of analysis:

1. General damages of $10,000,000 and Aggravated damages of $3,000,000

2. Special damages of $3,000,000

3. Punitive damages of $2,000,000

[132] I will deal with each category in order.

General Damages and Aggravated Damages

[133] General damages and aggravated damages often overlap in defamation cases because the role of compensation for loss of reputation encompasses personal mental distress, restoration of standing in a community, and the prolongation of damage during litigation arising from a failure to retract or apologize: Soliman v. Bordman, 2021 ONSC 7023, at paras. 198-99. A nominal amount is insufficient for vindication: Second Cup Ltd. v. Eftoda, 2006 CanLII 26174 (ON SC), at para. 43. Unlike bodily injury cases, there is no cap on damages. This does not mean defamation suits are gold mines. The reason for the absence of a cap is that there has been no great risk of award inflation, given the average award in the 1990’s in the order of $30,000: Hill, at para. 169.

[134] One reason for awarding significant amounts of general and aggravated damages in some defamation cases is the difficulty in proving special damages arising from the publication of words: Emeny v. Tomaszewski, 2019 ONSC 3298, at para. 44. There is a necessary balance between the non-economic and economic damages awarded in defamation cases. Here, the plaintiff had some ability to prove special damages and the effect on his reputation was not “at large.” Defamation cases are so fact-specific that little is gained by comparing individual cases: Soliman v. Bordman, 2021 ONSC 7023, at para. 208.

[135] Unlike a news broadcast or a press conference, the libel was communicated to four persons: the BCPIO’s SIFT entry officer and the three prospective employers. Obviously, these communications had a profound impact on the plaintiff’s career prospects in the banking industry. No one outside of the persons interacting with the system and the defendant during the month of May 2012 would have received the information accusing the plaintiff of complicity in fraud. That distinguishes this case from those in which false allegations of criminal or quasi-criminal behaviour have been published widely against professional persons in a context where integrity and trust are important to self-identity.

[136] Despite the narrow distribution of the defamation, I acknowledge the fact that it has been eating away at the plaintiff for a dozen years while he waited for the court to recognize that the libel was unsupported by any real evidence. Indeed, the bank advanced the defence of justification based wholly on hearsay. The toll of the libel and the litigation were clearly in evidence during the trial, in the contrast between the highly capable salesperson that Conway clearly respected prior to 2012 and the sometimes manic, sometimes child-like bearing of the plaintiff from beginning to end. For more than a decade, he has had to suffer the burden of explanation to his relations and to his community while his adversary sat on its hands and waited for the law’s delays to take effect.

[137] In recognition of these countervailing factors, one minimizing and the other worsening the effects of the libel, I assess general damages in the amount of $75,000 and aggravated damages in the amount of $100,000. The general damages amount reflects the immediate impact of the defamation on the plaintiff’s life in suddenly finding himself unemployable. The aggravated damages reflects the ongoing effect of the defamation beyond the economic consequences to be reflected in the pecuniary or special damages.

[138] Before concluding this section, I have also considered that general damages also encompass undefined economic losses such as earning capacity. The plaintiff claimed post-trial economic damage going forward until the age of 65. For the reasons stated below in the section regarding special damages, I conclude that the direct economic effects of the libel stopped in May 2019, upon the expungement of the SIFT alert. Thereafter, some indirect economic losses would have ensued, but the effect would have gradually mitigated after 2019.

[139] While it is true that if the plaintiff had rejoined the institutional banking industry at that time, he would have suffered a significant loss of opportunity in terms of the career differential between a 19-year veteran and an older employee, there was insufficient evidence to fill in what is essentially an imponderable question. Had the plaintiff been represented by a lawyer, I expect the lawyer would have hired a vocational expert and economist or actuary to fill that lacuna in the evidence. The court can make procedural allowances for self-represented parties, but it is not the role of the judiciary to base damages awards based on imagined evidence. The plaintiff adduced insufficient evidence of future economic harm beyond the expiration of the SIFT alert, such as denied employment applications, to prove such lasting effect beyond the principle that general damages can compensate for unspecified economic harm.

Special Damages

[140] The plaintiff submitted a spreadsheet tabulating his income prior to the libel and afterward. After being excluded from the mainstream banking sector, he has had to make a living as a mortgage broker. He compiled his income figures for a nine-year period before his resignation between 2002 and 2011. He then compiled his income for the period between 2012 and 2023 as a mortgage broker at Mortgage Alliance. The 2012 year included his first-quarter income while he was in the employ of the bank. He then projected his earning capacity for that second period based on the bank’s profit performance from secured lending during the same period. He rationalized this earning capacity calculation by pointing to real estate values in the Greater Toronto Area leading to larger mortgages and commissions.

[141] The defendant conceded that the actual income figures were accurate transpositions from income documentation produced by the plaintiff in the litigation and attached to his trial affidavit. The defendant took issue with the income loss theory in two principal respects.

[142] First, the income projections based on bank performance are not based on relevant evidence. The bank’s profits are based on a host of inputs, and one cannot interpolate an individual’s expected performance by assuming the commissions of an entire department, in turn postulated from an entire division of the bank’s holdings and enterprise. While to some extent the court might speculate that commission-sales bank employees do better in a rising market and do more poorly in a declining one, I agree with the bank’s general objection that the evidence is rather thin and based mostly on conjecture.

[143] Second, the defendant argued that the plaintiff resigned instead of seeing out the result of the investigation. In that sense, the plaintiff made it easier for the defendant’s employees to believe it could defame him. Had he not resigned, it would have been harder to dismiss him for cause than Roberts stated in her report. The defence built on this argument the fact that very little of the book of business Curtis had built up would have migrated with him to a new bank. Because of the commissioned nature of the employee compensation, he would have had to start from the bottom and work his way back to the top of the leaderboard.

[144] Thus, the defence argued, the plaintiff’s spreadsheet showed the plaintiff to have been enterprising enough to earn as a mortgage broker what he would have earned as an MDM at a new bank.

[145] I have also looked closely at the plaintiff’s income figures for the pre-resignation period. He included only ten out of twelve years at the bank. I can only surmise that he did not include the first two years because they would have brought down his annual average. His income during those ten years varied from $138,204 to $322,682. There was no pattern of growth. Rather, they were hills and valleys. The lowest income during that decade was in 2009, which I could guess had something to do with the 2008 meltdown from the sub-prime debt crisis of which Conroy testified. I am prepared to work with the average of $214,418, as a base. However, it is a stretch to say that Curtis could expect significant income growth in the post-2012 period. There were no trend lines or other indicia from which the court could be confident of Curtis’ income prospects.

[146] The plaintiff’s spreadsheet also contained an extended period after 2012 in which the plaintiff’s income dropped sharply. He argued that he was too busy litigating with an aggressive defendant to be earning money during that time. The defendant correctly pointed out the tasks of representing himself in the litigation, or even the aggressive legal tactics alleged by the plaintiff, could not have accounted for the periods of relative unemployment reflected in his income statements. Nevertheless, the court cannot ignore how the litigation process impacts litigants, and it is one recognized factor in the assessment of damages: Hill, para. 182.

[147] I have already considered the effect of the libel and the litigation on the plaintiff’s emotional condition and personality. Twelve years of litigation, including ten in this action, evidently took its toll on the plaintiff. The courts too frequently fail to consider the effect of litigation on the psyche and personalities of individuals, especially self-represented litigants. No one, even the ordinary denizens of courthouses and law offices, are immune from the effects of continuous exposure to conflict.

[148] Earlier this year, the Ontario Bar Association introduced a new award for legal-sector mental health initiatives, named in honour of its past-president and long-time advocate of justice sector mental health, Orlando Da Silva. Lawyers and judges now have access to assistance, but individual litigants do not. They must soldier on in an unfamiliar setting populated by rules and terminology they do not have the training to understand. Conflict exerts a depressive force on mental health, and the recognition of this effect is very much part of the experience and wisdom that the public expects from the judiciary when hearing cases where an individual’s personal identity or reputation is in issue.

[149] Conroy’s account of his former employee was tinged with no small amount of regret that Curtis had to be investigated and that he resigned. He had enjoyed travelling on at least one mortgage department foreign trip with Curtis. During the trial, the plaintiff was hot-headed, obsessed to the point of disruption with timekeeping by the trial judge, and fixated on chasing shadows. I cannot imagine that the person I encountered at the trial could have been as successful in selling mortgages and bringing clients to the bank as he evidently had been during the twelve years he worked at the bank. Similarly, his work as a junior retail bank officer required people skills and empathy.

[150] In weighing these aspects of the plaintiff’s case, including assessing his evidence, the court must be sensitive to the fact that the plaintiff qua witness at trial is likely a different person than the bank’s employee before the events of 2012. The court must also restrain itself from engaging in amateur psychology. The only evidence of the effects of the events and the litigation on Curtis was a terse family doctor’s letter stating his patient now requires medication for diabetes. It contained no causal nexus to anything related to the case and was certainly not a medico-legal report qualifying its use as a report pursuant to s. 52 of the Evidence Act, R.S.O. 1990, c. E.23. Beyond the testimony of the plaintiff himself, the evidence of the type of individual he was prior to 2012 came from Conroy himself. Conroy conceded that Curtis had been one of his best employees. He had nothing negative to say about his regard for Curtis prior to the 2012 investigation.

[151] The rough-hewn result must be that the court must take the litigation into account as a debilitating force in a self-represented litigant’s life, in that it is often a necessary consequence of defamation. During years of financial hardship between 2018 and the present, the plaintiff relied on handouts and loans from family and relatives. This dependency must, in turn, result in emotional and mental depression. I do not, however, accept the plaintiff’s facile evidence that the litigation or the bank’s conduct of the case was entirely responsible for his drop in income after he left the defendant in 2012. There was no real evidence of the plaintiff’s efforts to mitigate his economic damages after the initial attempts to secure employment at rival banks in May 2012.

[152] The starting point for the plaintiff’s past economic loss must be his resignation. Had the bank wrongfully dismissed him for cause (as they said they would have done) and then defamed him in the way they did, there could have been a contractual basis for starting the reference salary at as high as $288,552, his 2011 commission income. Because he resigned, the reference must be that of an MDM at another bank starting without an established customer base beyond the contacts in his smartphone. Even then, he would have to pry the business away from BNS.

[153] The plaintiff was an enterprising individual who, based on his historical record, one could imagine having been both charming and persistent. The defendant conceded in argument that he likely would have started at around where he started as a mortgage broker, in the range of $100,000 per year. I do not accept the argument that working for a rival bank would have tracked his mortgage broker income, since the $200,000 income from 2002 and 2003 as a junior MDM was already double that. I could also consider the value of various pension and benefits advantages of being a bank employee compared to an independent mortgage broker, but the plaintiff did not tender any evidence for comparison.

[154] All remedial damages assessments are discretionary: Jane Doe v. Canada (Attorney General), 2018 FCA 183, at para. 11. Special damages traditionally required strict pleading and proof, but the courts have relaxed this once inflexible rule to adapt to circumstances of a particular case: Reid v. Joy, 1999 CanLII 19035 (NL CA), at para. 75. The plaintiff could have obtained income figures for similarly ranked MDM’s for the relevant period, on examination for discovery. The plaintiff pursued interlocutory production orders for bank records such as “Toronto Region Delinquency Reports for Barry Ray” to show the plaintiff as a lower risk to the bank than his supervisory MDM, but none of this focused on an effort to prove his economic loss.

[155] It is not the court’s role to speculate about the evidence that the plaintiff could have tendered, be it the performance of his colleagues in the post-2012 period or some kind of labour market expert evidence. Perhaps it would have been more productive for the plaintiff to have researched the law of evidence and the law of damages than they label of esoteric torts that rarely succeed in the courts.

[156] Having reviewed the plaintiff’s damages calculations in depth, the best I can conclude is that the defendant’s libel and aggressive defence of the libel suit was partly responsible for the plaintiff’s loss of earnings being at a twelve-year average of $51,763. However, there were many factors, including litigation in other forums of other issues that had nothing to do with the libel. There was no evidence of market conditions in the mortgage broker industry compared to bank MDM’s during the Covid-19 years.

[157] Ultimately, the resignation and the exclusion from work at other banks makes the job at another bank the reference point and not a continuation of employment at BNS. Therefore, I will set the reference point for such employment at another bank at $150,000, growing to an average of $200,000 by 2019 when the SIFT alert expired. It is essentially an extrapolation of the twelve years between 2000 to 2012. I draw an adverse inference from the plaintiff’s failure to include the 2000 and 2001 figures in the calculation. I then take the figures from 2012 to 2015 before the significant decline in his income as a mortgage broker and draw the inference that his ability to mitigate the damages would have been capped at an average of $150,000 throughout the whole period. The $50,000 per annum differential translates to an amount of $350,000 over the course of seven years for the duration of the SIFT alert.

[158] The gap would have attenuated beyond 2019. The plaintiff had already reached the summit of his MDM group in 2011, whereas there was no evidence of his potential in the mortgage brokerage field. For the five remaining years, I calculate the differential to start declining from $50,000 to zero, meaning the special damages would amount to $125,000 for that interval.

[159] Bringing the two figures together, I assess the plaintiff’s special damages for loss of income at $475,000.

Punitive Damages

[160] The plaintiff’s claim for punitive damages is driven by the animosity and outrage he felt in the aftermath of the bank’s libel, including its failure to remove the SIFT alert until it expired in 2019. Angry plaintiffs claim punitive damages. The plaintiff is an angry individual. The impact on the plaintiff of the libel and subsequent failure to retract it is the subject of general and aggravated damages. Punitive damages, however, arise from a different and non-compensatory rationale. Plaintiffs need to abstract themselves from their anger and consider the public policy role of punitive damages. As described in Hill, at para. 196:

Punitive damages may be awarded in situations where the defendant's misconduct is so malicious, oppressive and high‑handed that it offends the court's sense of decency. Punitive damages bear no relation to what the plaintiff should receive by way of compensation. Their aim is not to compensate the plaintiff, but rather to punish the defendant. It is the means by which the jury or judge expresses its outrage at the egregious conduct of the defendant. They are in the nature of a fine which is meant to act as a deterrent to the defendant and to others from acting in this manner. It is important to emphasize that punitive damages should only be awarded in those circumstances where the combined award of general and aggravated damages would be insufficient to achieve the goal of punishment and deterrence.

[161] Institutional defendants such as the one in this case rely on the rule of law and the civil remedies available through the courts to support its business. Indeed, mortgage remedies and other facilities afforded by the legal system are the backbone of the financial services sector. When such a defendant is sued by an individual litigant for a clear wrong, the bank seeks the shelter of the procedural rules of same legal system to shield it against accountability for its wrongs. The quantum of punitive damages can also depend on the defendant’s means to pay them: Noreast Electronics Co. v. Danis, 2018 ONSC 5169, at para. 161.

[162] In Whiten v. Pilot Insurance Co., 2002 SCC 18, [2002] 1 SCR 595, the Supreme Court upheld a jury award of $1,000,000 in punitive damages for the conduct of a homeowner’s insurer in denying coverage based on an unfounded allegation of arson. The analogy to unfounded criminal allegations is not lost on this court. Moreover, leaving Curtis without a livelihood in the immediate term was not too different from leaving a family without of home. However, upholding a jury award is not the same as setting a range for damages, because of the issues relating to the standard of review. Undoubtedly, the Supreme Court judges would have awarded lower amounts had they been trial judges.

[163] While I do not consider the flawed process of the bank’s investigation of Curtis as warranting punitive remedies from the court, I cannot say the same about the bank’s failure or refusal to lift the alert once it ought to have realized that the original defamation was no longer supportable by any means. The bank ought to have taken a hard look at the evidence and the law of defamation to reach the conclusion that the SIFT alert had to be lifted. Although this was not an employment law case, as I had to remind both parties repeatedly during the trial, defamation that impedes future employment is particularly harmful because of the importance our society places on employment on personal identity and social standing. The bank did not introduce any evidence of investigation of Curtis after the interview. The only information that followed it was a note that one of the mortgage brokers had been on the exclusion list of another part of the mortgage department. Somehow, this information to which Curtis and his colleagues had not been privy became part of the bank’s judgment regarding Curtis’ involvement in the frauds.

[164] The bank’s actions are reminiscent of R. v. Fagan, [1969] 1 Q.B. 439; R. v. Lajoie, 1989 CanLII 7225, 47 CCC (3d) 380 (QC CA), at 383 CCC, in which the motorist’s inadvertent treading on the policeman’s foot became assault after the failure to move the car off it. In that instance, there was nothing unlawful about the accident. However, the failure to relieve the pressure on the foot amount to an intentional act warranting criminal sanction. When a libel comes to the attention of the utterer as defamatory, as in the case of a libel notice to a newspaper, the publisher can then dial down the effect of it by investigating and taking remedial measures: Libel and Slander Act, R.S.O. 1990, c. L.12, s. 5. Moreover, early attempts at resolution are admissible as evidence of mitigation by the defendant, pursuant to s. 10.

[165] No libel notice was introduced in evidence. However, the 2014 statement of claim certainly constituted such notice, and the libel continued to be published and have a prejudicial effect until May 2019. The defendant could have taken a harder look at the evidence and concluded that the SIFT alert was unjustified. Or, if it was justified, it could have marshalled and preserved the evidence to avoid the finding by the court that there was insufficient evidence of justification. Indeed, the case at trial heard of evidence that could have been material to the issues, such as the written policy circumscribing MDM’s dealings with realtors who were mortgage brokers and the audio recording of the Curtis interview, which existed at some point but lost either forever or until the very eve of trial. While the impact such evidence could have had on the trial is unknown, the fact that the bank failed to preserve it while the libel remained published for seven years demonstrated a profound indifference to whether the libel was, in fact, justified. This should not be the conduct of a Canadian institution that depends on the rule of law and the courts to uphold its existence and ability to earn profits for its shareholders.

[166] The failure to investigate the justification of the libel over the course of several years is already the subject of the combined general and aggravated damages. However, there must be some award of punitive damages that recognizes the obligation of a defamer to correct or mitigate the harm beyond the compensatory award.

[167] I therefore award punitive damages against the bank in the amount of $200,000 for the fact that it never once considered revisiting the original flawed investigation to determine whether there was proof of Curtis’ involvement in the shelter fraud perpetrated by the bank’s mortgage applicants and customers. Too frequently, institutional defendants treat litigation as a cloak for freezing activity. The court must make this absolutely clear: litigation is the time to investigate, investigate, and investigate. Doing nothing and expecting the lawyer to pull the rabbit out of the hat is unfair to the lawyer, risks defeat at trial, and forces the opposing party to endure unnecessary litigation.

PREJUDGMENT INTEREST AND TOTAL DAMAGES

[168] The action was commenced on April 22, 2014. The prejudgment interest rate is therefore 1.3%., pursuant to the published tables under s. 127 of the CJA.

[169] The base date for the accrual of prejudgment interest cannot be determined precisely, because the initial SIFT alert was a “pull” publication as opposed to a “pushed” broadcast. This meant the libel originated when prospective employers made their database inquiries and were then told by the defendant that the plaintiff was reported to SIFT because he had submitted fraudulent documentation. Assuming the plaintiff would have started employment with a different bank shortly after he left the defendant, I will set June 1, 2012, as the starting date for prejudgment interest.

[170] On the $175,000 in general and aggravated damages, I calculate 11.9 years of interest to amount to $27,072.

[171] On the $475,000 in special damages, I calculate the prejudgment interest differently for the first seven years of damages and the last five years. Seven years of prejudgment interest on $350,000 amounts to $31,850. Dividing that in half due to the accrual calculation results in a figure of $15,920. In the ensuring 4.9 years, the interest is $22,295. The interest on the remaining $125,000 of special damages operates on an inverse accrual basis over 4.9 years, amounting to $3,981. The total of the prejudgment interest on special damages therefore amounts to $42,196.

[172] Pursuant to s. 128(4)(a) of the CJA, there is no prejudgment interest on punitive or exemplary damages of $200,000.

[173] I therefore find that the defendant is liable to the plaintiff to pay a total of $850,000 in damages, plus pre-judgment interest in the amount of $69,268.

COSTS

[174] The plaintiff is presumptively entitled to costs of the action, including costs for his time as a self-represented party. If there are true rule 49 offers that may affect the costs outcome, the parties should include them in their cost’s submissions. If the parties do not settle the costs, the plaintiff may serve and file a bill of costs and costs submission on or by May 1, 2024. The defendant may then file costs submissions on or by May 15, 2024. Once the costs materials have been served and filed, the civil intake office is hereby directed to send them to my judicial assistant.

[175] It was evident from the trial record and the conduct of the trial that many of the steps and delays have been occasioned by the plaintiff’s belief in the benefits of pleading esoteric torts and prolix arguments. Accordingly, the plaintiff is directed to exclude from the bill of costs any costs associated with pleadings motions. As I stated in my trial management endorsement of February 23, 2024, the plaintiff also prepared a multi-volume set of documents contrary to the court’s practice direction that the judge will only have access to materials uploaded to CaseLines. He will therefore not be reimbursed by the defendant for the cost of photocopying and binding those materials.

Akazaki J.

Released: April 19, 2024