Mr M Onyekwere v Trilateral Research Ltd: 2200391/2024
EMPLOYMENT TRIBUNALS
Case No 2200391/2024
Between
Mr M OnyekwereClaimantTrilateral Research LtdRespondent
Before
Employment Judge B SmithIn person for claimantDate 15 November 2024
REASONS
[1]The claimant applied for jobs at the respondent, a company which operates in the field of AI, social research, and consulting services in the fields of data protection, cyber-risk management, ethics innovation, and responsible AI.[2]ACAS conciliation commenced on 24 November 2023 and concluded on 22 December 2023. The claim was presented on 7 January 2024.[3]The claimant brings a single claim of direct race discrimination (s.13 Equality Act 2010 ‘EQA’). The issues in the claims include about whether the claimant was not invited for interview, and his job applications rejected, because of his race. B Procedure, documents, and evidence heard[4]The respondent was represented by Ms Douglass, a HR consultant. The claimant represented himself throughout. The Tribunal took this into account and took all reasonable steps to ensure that the parties were on an equal footing. This included neutral questions of the witnesses where appropriate. It was also necessary, for fairness, for the Tribunal to interrupt the claimant’s cross-examination on several occasions in order to ensure that witnesses were only cross-examined on an accurate basis of their previous evidence and with an accurate description of the other evidence in the case. Also, for fairness, it was necessary to interrupt one element of the claimant’s cross-examination of Ms Finn. This was because she was being cross-examined on the basis that certain documentation was not included in the hearing bundle. However, we were satisfied upon reading inter-party correspondence and after hearing from the parties that the reason why that material was not included in the hearing bundle was due to the objections of the claimant about late service. This was in relation to the extent to which the relevant decision makers had undertaken unconscious bias training. This was not a central point to the claim, but the Tribunal considered it unfair for the witness to be cross-examined on the basis that there was no documentary evidence in support of her contentions when as a matter of fact that material did exist but had not been included in the Tribunal bundle.[5]The Tribunal also rephrased some questions by both representatives to ensure that they were relevant to the issues in the case and to ensure that the respective positions were put to the relevant witnesses.[6]The Tribunal also restricted questioning due to relevance, and imposed time limits where necessary to further the overriding objective. The parties were given notice of time limits on cross-examination and the Tribunal was satisfied that both parties were given a full opportunity to challenge the evidence as necessary for a fair hearing.[7]The Tribunal also gave a brief and neutral summary of some relevant legal points at the start of the hearing and before submissions. Timetabling during the hearing was with the agreement of the parties.[8]No adjustments were required or asked for by any of the parties or witnesses. Regular breaks were taken throughout.[9]The claimant gave evidence under affirmation. The respondent’s witnesses were Ms Alderton, Ms Finn, Ms Battersby, and Ms Sveinsdottir. All gave evidence under affirmation.[10]The list of issues was set by order of EJ Emery dated 19 April 2024. They were agreed by the parties at the start of the hearing. The parties agreed that no applications to amend the claims or response were required at the start of the hearing.[11]The issues for liability were: 2. Direct race discrimination (Equality Act 2010 section 13)2.1 Did the respondent do the following things: 2.1.1 Reject the claimant’s job applications -2.1.1.1 April 20232.1.1.24 July 2023 (if made) and2.1.1.3 September 2023?2.2 Was that less favourable treatment? The Tribunal will decide whether the claimant was treated worse than someone else was treated. There must be no material difference between their circumstances and the claimant’s. If there was nobody in the same circumstances as the claimant, the Tribunal will decide whether they were treated worse than someone else would have been treated – a ‘hypothetical comparator’. The claimant relies on the job-applicant ‘Michael O’Reilly’ as a hypothetical comparator.2.3 If so, was it because of race?[12]Separate issues for remedy can be found in EJ Emery’s order. Issue 1 related to time limits.[13]The respondent confirmed that it no longer took issue with time limits. The Tribunal therefore proceeded on the basis that the claims were in time because the relevant conduct was a continuing act with the last act within time.[14]The agreed documents were:(i) Hearing bundle paginated to 279;(ii) Claimant witness statement (‘Written address’);(iii) Respondent witness statement bundle (31 pages in pdf);(iv) Agreed chronology;(v) Agreed cast list; and(vi) Key documents list.[15]At the Tribunal’s request, in light of the witness evidence, we also required that the following documents be sent to the Tribunal:a. ‘Trilateral Phase 3 Doc 1’ which was an attachment to an email dated 5 December 2023 from the claimant to the respondent; andb. Interparty correspondence dated 20 August 2024 (3 emails).[16]The Tribunal only took into account those documents which the parties referred to during the course of the hearing in accordance with the normal practice of the Employment Tribunals. The parties were made aware of this from the outset.[17]Both parties made written submissions. The respondent made oral submissions in reply. The claimant did not make oral submissions, although he was given an opportunity to do so.[18]Following submissions the parties confirmed that no issues arose about procedural unfairness during the hearing. C Relevant Law (i) Burden of proof in EQA claims[19]The burden of proof for the EQA claims is governed by s.136 EQA:(1) This section applies to any proceedings relating to a contravention of this Act.(2) If there are facts from which the court could decide, in the absence of any explanation, that a person (A) contravened the provision concerned, the court must hold that the contravention occurred. […][20]It was held in Field v Steve Pie [2022] EAT 68 at [37]: ‘In some cases there may be no evidence to suggest the possibility of discrimination, in which case the burden of proof may have nothing to add. However, if there is evidence that discrimination may have occurred it cannot be ignored. The burden of proof can be an important tool in determining such claims. These propositions are clear from the following well established authorities.’ Further at [41] that ‘if there is evidence that could realistically suggest that there was discrimination it is not appropriate to just add that evidence into the balance and then conduct an overall assessment, on the balance of probabilities, and make a positive finding that there was a non-discriminatory reason for the treatment.’[21]It is not sufficient for the employee to only prove a difference in protected characteristic and a difference in treatment in order to shift the burden of proof: Madarassy v Nomura International Plc [2007] EWCA Civ 33.[22]An inconsistent explanation for a difference in treatment may be a factor which shifts the burden: Veolia Environmental Services UK v Mr M Gumbs UKEAT/0487/12/BA.[23]Once the burden has shifted, the employer must prove that less favourable treatment was in no sense whatsoever because of the protected characteristic: Wong v Igen Ltd [005] EWCA Civ 142. (ii) Direct race discrimination[24]Direct discrimination is prohibited conduct under s.13 EQA: A person(a) (A) discriminates against another(b) (B) if, because of a protected characteristic, A treats B less favourably than A treats or would treat others. […][25]The comparator’s circumstances must be the same as the claimant’s, or at least not materially different. This is because s.23 EQA says: (1) On a comparison of cases for the purposes of section 13 … there must be no material difference between the circumstances relating to each case. […][26]The protected characteristic need not be the only reason for the less favourable treatment, or the main reason: London Borough of Islington v Ladele [2009] IRLR 154 (EAT). The decision must be more than trivially influenced by the protected characteristic.[27]The question of less favourable treatment can be intertwined with the reason for that treatment: the principal question is why was the claimant treated as he was? If there were discriminatory grounds for that treatment then ‘usually be no difficulty in deciding whether the treatment …was less favourable than was or would have been afforded to others.’ There is a single question: did the complainant, because of a protected characteristic, receive less favourable treatment than others’: Shamoon v Chief Constable of the Royal Ulster Constabulary 2003 ICR 337 HL.[28]Also, in Stockton on Tees Borough Council v Aylott 2010 ICR 1278, CA, Lord Justice Mummery stated: ‘I think that the decision whether the claimant was treated less favourably than a hypothetical employee of the council is intertwined with identifying the ground on which the claimant was dismissed. If it was on the ground of disability, then it is likely that he was treated less favourably than the hypothetical comparator not having the particular disability would have been treated in the same relevant circumstances. The finding of the reason for his dismissal supplies the answer to the question whether he received less favourable treatment’.[29]Where the question is addressed in this order the Tribunal need not necessarily identify the precise characteristics of the hypothetical comparator: Law Society and ors v Bahl 2003 IRLR 640 EAT.[30]Where a job application is not a genuine one, this may be a reason to find that there is no less favourable treatment: Keane v Investigo and others UKEAT/0389/09/SM. D Findings of fact[31]We have only resolved the issues of fact necessary to make our decisions. Findings of fact are made on the balance of probabilities. We confirm the cast list as findings of fact. We only explain our reasons for making these findings where there was a material dispute between the parties. The vast majority of findings are made based on the documentary evidence included in the hearing bundle or unchallenged witness evidence.[32]There is no dispute about the authenticity of the documents. (i) Background[33]The claimant has a background which includes data protection and compliance, including from the financial services sector. The claimant uses two email addresses. One is referred to in this decision as his ‘85’ email address. The other is referred to as his underscore address.[34]The respondent undertakes various activities as set out in the introduction above. These include providing data protection officer-type services to the company’s clients. The respondent is a small company with about 70 people employed in the UK, 25 in Ireland and various consultants in Europe.[35]On 8 March 2023 the relevant people at the respondent undertook training which included unconscious bias training. This included those who made the reviewing decisions about the claimant’s CVs later on. We make this finding because we accept the evidence of the respondent’s witnesses that this happened. Whilst the claimant relies on a lack of documentary evidence in support, Ms Finn in particular gave convincing and credible oral evidence about the training, including remembering being late to attend a child’s sports event as a result of that training. Also, we note that the claimant had objected to documentary evidence being included in the hearing bundle by the respondent (on the grounds that it was provided after the relevant Tribunal deadlines), although we understand that he later conceded that it could be included as disputed documentation. In the circumstances, there is no good reason not to accept the oral evidence about the training that took place. (ii) The July 2022 application[36]In July 2022 the claimant made an application for a Senior Research Analyst Climate/Legal (‘Research Analyst’) position at the respondent. This was a materially different role to later applications he made for a Data Protection Adviser role. The claimant used the underscore address to make this application. We make this finding because it is clear from the respondent’s recruitment records system which confirmed the claimant’s application to that email address.[37]The claimant was rejected for this role. The claimant was rejected for this role because he did not have the right skillset. We make this finding because the respondent’s recruitment systems records state this as the reason. Also, the claimant accepted in evidence that he did not have the right skillset for that particular role. We find that the claimant was informed of his rejection for that role by email. Although the claimant does not have any memory of him receiving a rejection communication from the respondent about that role, the respondent’s recruitment system has a record dated 23 January 2023 suggesting that an email rejecting the claimant was sent. Although we accept that on the face of the record a generic and potentially incomplete email was sent, on balance we consider that the system records are likely to be accurate. The reason included in the rejection email included that the respondent had ‘ended up moving forward with another applicant’. We do not consider the claimant’s memory about whether or not he had received that rejection to be sufficient to find that no rejection was sent in light of the system records. We had no reason to doubt the accuracy of this particular record. Also, the claimant accepted in relation to his later applications that it was possible that rejections had been sent and he did not see them or keep them. (iii) The applicant assessment process[38]We accepted the evidence of the respondent’s witnesses as to the shortlisting process undertaken for the relevant job applications subject to this claim. This is because there is no good reason to doubt their evidence. Whilst it was not expressly supported by documentary evidence it was not undermined by any other evidence. Also, it was consistent with what documentary evidence existed. Also, the process (as described) would not necessarily have generated any more documentary evidence than we were provided with.[39]The process was as follows. Applicants’ CVs and cover letters were reviewed by cluster leads within the relevant team. It was not recorded which cluster leads reviewed which CVs. There was no formal or documented marking process. In general, if a candidate met the essential criteria than they may be shortlisted for interview. However, cluster leads were permitted to take into account any particular client need that they had in mind at the time outside of the formal essential criteria. They were also permitted to compare the applicant to other CVs and take into account the strength of the application in terms of the criteria. It was therefore an evaluative rather than automatic process. For jobs where there were a high number of applicants, as was the case for the relevant role subject to these claims, a further informal shortlisting process would take place within any first pool of candidates not initially rejected. The reviews were always undertaken by two cluster leads, and generally they were looking at CVs on the same screen. Cluster leads had many other aspects to their roles and this, combined with the high volume of applicants, meant that each CV was not necessarily reviewed for a lengthy period of time. The decisions about who would be rejected (also referred to as disqualified) were made by the cluster leads and, for this particular exercise, those decisions would be reflected on the system by the administrator. These findings are expressly supported by the most relevant evidence of Ms Finn, who was the director of the relevant unit, and Ms Sveinsdottir, who was one of the cluster leads.[40]We find that the approach used by the respondent was flexible and took into account the particular ultimate client that the reviewer had in mind for any particular applicant. The ultimate client changed over time and between reviewers. In any event, the fact of rejection did not mean that the essential criteria were not in fact met by the applicant. The fact of rejection just meant that the applicant was not shortlisted for interview. On the basis of the above, some candidates may have met the essential criteria but not ultimately be shortlisted for interview. This is because other candidates were stronger. This was not reflected in any formal marking on the system.[41]We consider this system to be proportionate to the nature and type of role in question.[42]For those reasons we reject the claimant’s assertion that the respondent’s approach was that if the essential criteria were met, he would be shortlisted for interview. The claimant has taken small sections of the respondent’s witness evidence in isolation to suggest that this was the case, but this disregards the rest of the witnesses’ written and oral evidence, which was clear. The process was considerably more flexible and dynamic than the claimant sought to present. Ultimately, the respondent’s witnesses were better placed than the claimant to describe what the processes were, and we have no reason to doubt their evidence. (iv) The April 2023 application[43]The claimant applied for a further role at the respondent by cover letter and CV in April 2023. The cover letter and CV were as included in the bundle. The role was for Data Protection Adviser (referred to in the chronology as Data Protection Officer) and the advert was as set out in the bundle.[44]The respondent advertised for this role throughout the relevant period, which included April 2023 to around September 2023. Although a single role is referred to, in fact this was a rolling recruitment exercise and more than one person could be hired. The job description did not change. However, the applications were assessed with the clients of the respondent in mind. The ultimate clients changed over time. We make this finding because we accept the respondent witnesses oral evidence on this point, in particular Ms Finn and Ms Sveinsdottir which was not materially undermined by any cross-examination, documentation, or other evidence. The applications were first assessed by cluster leads in the relevant team at the respondent. These cluster leads had particular clients in mind when assessing applications. For example, Ms Sveinsdottir was looking for the role to fulfil a particular EU institutional client’s needs in around April 2023. For the August 2023 period she was looking to recruit in respect of a UK-based client. Other cluster leads were looking for different clients over that period. It was the case, therefore, that although the advertised job description remained the same, the respondent was looking for slightly different applicant backgrounds, or emphasis in their skillset and experience, at different times. For example, we accept Ms Sveinsdottir’s evidence that around April 2023 when she was looking in respect of an EU institutional client, she was looking for someone with pure GDPR-type experience as opposed to someone whose compliance experience was split between GDPR and other areas such as anti-money laundering (‘AML’) and financial services compliance (know you client, or ‘KYC’). She also would have preferred EU experience at that time. For August, she was happy to see a broader range of experience for the particular ultimate client she had in mind who was UK-based. There was no good reason to doubt her evidence on this point, and we make those findings of fact accordingly.[45]The claimant’s April 2023 CV included under ‘Objective’: ‘Highly experienced lawyer, compliance and data protection professional seeking a challenging role that leverages my extensive knowledge in financial services compliance, AML/KYC, GDPR and risk management to drive organizational success and ensure regulatory compliance.’[46]The professional summary included ‘Over 5 years of experience in legal, compliance and data protection roles, with a proven track record in AML/KYC, GDPR and CDD implementation and management.’[47]Key skills were listed as follows: AML/KYC, GDPR, CDD, Risk management, regulatory compliance, Training and development, enhanced Due diligence, data protection.[48]The most recent professional experience was from May 2022 which included the following: - Developed and maintained a comprehensive risk framework, resulting in a 15% reduction in operational and regulatory risks across the UK and Ireland marks. - Streamlined the client onboarding process, leading to a 20% improvement in response time for client-facing inquiries. - Successfully implemented GDPR and AML/KYC training programmes, increasing staff awareness and adherence to regulatory requirements.[49]The claimant’s qualifications included LLB and LLM degrees within the jurisdiction of England and Wales and Barrister of Law at Nigerian Law School.[50]The claimant’s application was acknowledged by email on a date which is unclear. The application was made using the claimant’s 85 email address. We make this finding because this is the email the respondent’s recruitment system responded to. We note that the email address the respondent’s system responded to did not necessarily match the email address referred to in the separate document that was the claimant’s CV.[51]The claimant’s April 2023 application was rejected by the respondent on 16 May 2023, but the respondent included in its rejection that ‘We have reviewed your application and regret to inform you that yours was not selected for further consideration, however we will be advertising more positions in the coming months. We hope you will keep us in mind and encourage you to apply again’.[52]The April 2023 application was marked as ‘disqualified for lack of experience’ on the recruitment system. This was one of a number of drop down boxes on the system. We accept the respondent’s witness evidence that this reason was not necessary the full reason or the most accurate description of the reason why an applicant was rejected. This is because, as a result of the volume of applications, generic responses were sometimes provided as a bulk response to rejected applicants. That element of the system itself also only allowed one reason to be used for each applicant although it was possible for additional notes to be made. We do not find, however, that the use of bulk rejection reasons is of itself suspicious in the circumstances of this case. Also, this system record of a reason for rejection does not appear to have been communicated to the claimant given the wording used in the 16 May 2023 communication above. This is consistent with the internal records for rejection at that stage being generic in nature and that little can be inferred from the use of any particular dropdown marker.[53]The respondent recruited two individuals for this role. One was a white woman and one is described by the respondent as black, Asian, or minority ethnic. We make this finding because we accept the respondent’s witnesses oral evidence of this and there is no reason to doubt it.[54]We reject the claimant’s submission that it is more likely than not that his April application was rejected by the recruitment administrator acting alone. There is no clear evidence that this is what happened. It is mere speculation on the claimant’s part. The claimant relies on a lack of documentary evidence around the review of his CV and the fact that the administrator was later found, on the evidence of Ms Battersby, to be not always good at following procedure. However, this does not demonstrate the point the claimant wishes to make. This is because there is no evidence that the administrator made a mistake in respect of the April application, whereas Ms Battersby did personally see mistakes in relation to the administrator’s August conduct. Also, on the respondent witnesses evidence, which is not meaningfully undermined by anything, there is no reason why documents would have been produced about the initial review of CVs other than the fact that they were accepted, were shortlisted to interview, or rejected with a generic-type reason chosen from a drop-down box on the system. The respondent has provided consistent, credible and reliable oral evidence from multiple witnesses about the processes followed and there is no reason to doubt that evidence. Also, the Tribunal takes into account that the roles were not particularly senior by nature and we would not have expected significant amounts of documentary evidence to have been generated about the initial sift about who to take forward to interview in circumstances where there was a large volume of applications over a reasonably significant period of time.[55]We also accept the respondent witnesses’ clear evidence that the administrator was acting on the instructions of the cluster leads in relation to the sifting of applicants for this particular role given the number of applicants. There is no good reason to find otherwise. The July 2023 application[56]The claimant then applied for the similarly advertised position in July 2023. This was in his own name by CV only.[57]There is a dispute between the parties as to whether the July CV was materially different to the April CV. The claimant accepts that there were some wording changes but maintained that on the face of the CV he still met the respondent’s essential criteria and therefore should have been shortlisted. The claimant accepted that it was, however, an improved CV.[58]The July CV included the following, with the full version as set out in the bundle. Under objective the claimant stated: ‘Analytical and collaborative Nigerian qualified solicitor and compliance professional with over 5 years of experience in financial services. A selfmotivated team player, committed to ethics and innovation, with a strong background in GDPR, AML/KYC, and risk management.’ Key skills included the following: ‘• Data Protection & Security: Detail-oriented background in data protection, privacy, and security within the financial services sectors. Proficient in time management to meet tight SLAs and ensure strategic compliance.• Compliance Expertise: Proficient in GDPR, AML/KYC, and CDD which is directly transferable in the blockchain space. Strategic approach to risk management and continuous development.• Risk & Project Management: Effective risk management, process improvement, and demonstrated ability to manage projects analytically and implement strategies.• Stakeholder Engagement: Excellent ability to interface with and manage stakeholders through written communication, ensuring alignment with their specifications. Collaborative approach to service delivery and sales engagements.• Collaboration & Consulting: Ability to work collaboratively across functions, offering innovative solutions and consulting expertise in startup and small business environments.• Technical Proficiencies: Maintenance of data inventory, understanding of security protocols, adeptness in data flow applications, and prompt issues resolution.• Communication & Attention to Detail: Strong written and verbal communication skills and attention to detail in all aspects of work, from research to feedback. Skilled at translating complex regulatory status and updates via.’[59]The most recent professional experience was from May 2022 and stated the following: - Collaborated with consumer stakeholders in crafting a data privacy strategy for all group entities, focusing on market needs, startup growth, and the crypto landscape. - Analytically developed and implemented a risk framework, ensuring it adhered to set specifications and mitigated operational/regulatory risks by 15%. Employed strategic time management to meet SLAs. - Championed process improvements in client onboarding, enhancing data flow, strategic service delivery, and bolstering security through innovative designs.[60]In terms of the differences between the content of the April and July CVs, we agree with the respondent witnesses that there are difference between them. These differences are more in emphasis of the claimant’s skills and experience, and are also presentational in nature. This was the view of the respondent witnesses. We accept that there is a limitation to the respondent witnesses’ analysis of the claimant’s CVs because the only witness who might have reviewed it first time around was Ms Sveinsdottir, as a cluster lead, and exactly who reviewed the claimant’s CVs was and remains unknown, save for the fact that Ms Sveinsdottir did review the claimant’s later August CV. However, we had access to the documents and there are clear differences in the wording used.[61]Specifically, the claimant puts a different emphasis on his GDPR-type skills, in part through the order in which his skills are listed. GDPR is put in a more prominent position in the July CV. Also, in describing his most recent work experience, the claimant describes what he did in terms which were more focussed, and the words used carried more significance. For example, in the April CV the claimant ‘Developed and maintained a comprehensive risk framework, resulting in a 15% reduction in operational and regulatory risks across the UK and Ireland marks.’ However, in the July CV the claimant ‘Analytically developed and implemented a risk framework, ensuring it adhered to set specifications and mitigated operational/regulatory risks by 15%. Employed strategic time management to meet SLAs.’ We agree that implementing a risk framework is more significant experience than simply maintaining a risk framework. Also, the claimant accepted that the CV had been strengthened.[62]There was a further important element of dispute between the parties about the July application. This is because the respondent had no record of the application in its systems despite an extensive investigation having been carried out. However, we find that the claimant as a matter of fact did make the July application to the respondent, through its old recruitment system. We make this finding because the claimant has given oral evidence that he did and provided the Tribunal with the CV he used. He described to us the process he used to create the CV. Also, and most importantly, the claimant received an email from the respondent’s recruitment administrator which indicated that an application had been received and was being reviewed. This email was sent on 14 July 2023. That email was sent via the old recruitment system as opposed to from the administrator’s Trilateral email address. That email was sent to the claimant’s underscore email address which is the address he says was used to make the application. Also, we note that this is different to the 85 address which the claimant used for the April application. We therefore find that it is more likely than not that the administrator’s email is referring to the July application than the April application because it is to a different email address. There is no good reason to believe it was in response to the 2022 application. Also, the claimant had already been rejected for the April application and therefore this email is more consistent with the claimant having made the July application, as opposed to no July application ever having been made.[63]We also accept the claimant’s evidence that the July application was made by the claimant because he suspected that AI sifting was being done by the respondent and that improved wording of his skills and experience. Also, we accept that the July application was a genuine one. This is because there is no good reason to find otherwise.[64]We accept the respondent’s position and evidence that they have no record of the administrator’s email and have been unable to find it. We make this finding because there is no evidence which clearly undermines this and we heard oral evidence of extensive investigations having been carried out to locate the email. However, it was not in dispute between the parties that this email was in fact sent to the claimant from the respondent via the recruitment system. This was agreed by the respondent at an earlier case management hearing on 19 August 2024 (EJ Anthony). It is clear to us, on the evidence, that the respondent only has this email now as a result of the claimant sending a copy of it to the respondent in December 2023.[65]The claimant’s position was that the respondent’s position about the administrator email is evidence of a cover up. We reject that position. The narrative of events is that the claimant sent the respondent a copy of the email in December 2023. The respondent did not accept that email on its face as being necessarily authentic because they had no copy of it at their end and, on its face, the email did not contain the hallmarks of a genuine email. This was the evidence of Ms Alderton. A copy of that email was provided to the Tribunal and we agree, on the face of the email, that the pdf document sent to the respondent by the claimant in December 2023 is not necessarily, on its face, enough evidence to be clear that the email was real. On its face, that document could easily have been created in Word. However, on the day after a case management hearing in April 2024 the claimant sent the respondent a further document which contained the email’s raw data. The claimant’s representative then spent a significant period of time not on the case due to medical reasons. The claimant also sent the respondent an expert report about the email and applied to the Tribunal for permission to rely on that report. It then followed that the respondent accepted at a further case management hearing in August 2024 that, as a matter of fact, the email had been sent. However, it should be noted that there is no evidence, and it was expressly denied by the respondent, that this change of position was as a result of any information that the respondent had other than that provided by the claimant. The claimant submitted that this is evidence of a cover up. The claimant also submitted that the respondent would only have admitted that the July email had been sent if it in fact it knew that the email was genuine all along. We reject that submission. It does not follow as a matter of logic. We accept the respondent’s evidence that there was no cover up because there is no cogent evidence to undermine that evidence or find otherwise. The fact that a party changes its position after receipt of further information from the other party demonstrates nothing of consequence.[66]We also do not find that the claimant’s allegations of selective disclosure by the respondent are supported by the evidence. In particular, the claimant has confused disclosure, namely the provision of copies of documents actually held by the respondent (or, at least, stating to the claimant that such documents exist), with the position taken by the respondent in litigation about various facts. A change of position does not in of itself show that there has been a failure of disclosure.[67]The claimant submits that part of Ms Battersby’s witness statement suggests that there was material in the administrator’s Trilateral email account which was from the claimant. However, when taken in context with her oral evidence, we accept her evidence that the administrator’s Trilateral account was not linked to the recruitment systems email account. The claimant has not himself suggested that he corresponded with the administrator’s Trilateral account (as opposed to via the recruitment system). Ms Battersby accepted that a single part of her statement was perhaps poorly worded but was very clear in her oral evidence about the lack of material in the administrator’s Trilateral account. We have no reason to doubt her oral evidence and therefore do not do so.[68]We do not find that the claimant’s later email to the old recruitment system’s general email address ‘GDPR is coming for you’ on 2 April 2024 demonstrated that the respondent was wrong in respect of its inability to recover emails from the administrator’s email account. This only shows the continued operation of a generic email address which was not relevant to the claimant’s claim.[69]We also find that the respondent’s email to the claimant asking whether he wanted his data to be preserved in August 2023 is inconsistent with the respondent seeking to cover up, hide or intentionally delete material. Outcome of the July 2023 application[70]We find that the July application was never meaningfully reviewed by the respondent. This is because there is no evidence that this happened. We do not accept that the administrator’s July email suggests that there was a meaningful review of the CV because we heard an abundance of evidence about the use of generic and holding emails by the respondent. This is perfectly normal given that the respondent received high volumes of applications for these posts and they were initially reviewed by relatively senior members of teams with many other operational demands on their time.[71]We also consider that the lack of rejection or acceptance records about the July application is indicative of it never having been meaningfully reviewed. All of the claimant’s other applications were either progressed to interview or rejected and records exist of these outcomes. No such records exist for the July application.[72]It is slightly unclear why the July application was never meaningfully reviewed by the respondents. However, we accept the respondents’ witnesses evidence that they were unaware of the application for the reasons outlined above. We find that the claimant’s July application was not at any time specifically rejected by the respondent. This is because there is no evidence that this happened, such as a rejection email or record on the recruitment system as happened with all of the other applications made by the claimant.[73]We consider that it is more likely than not that the claimant’s July application and related emails were deleted in error by the respondent, or otherwise lost, in the transition between recruitment systems.[74]We make this finding because this is the most likely outcome based on all of the evidence. It is not in dispute that shortly after the claimant’s July application was submitted to the old recruitment system the respondent moved to a new recruitment system. This new system was used for the August application. The respondent sought to copy information from the old system to the new system but also sent an email to applicants, including the claimant, about whether or not they wanted their data to be preserved. The transfer was a significant task including review of around 4000 records. A significant amount of data was deleted around August 2023. At this time the deletion was being done by the recruitment administrator and Ms Battersby. Around this time the administrator had given notice and also took some leave. It was discovered by Ms Battersby that the administrator was meant to be creating logs of what material had been deleted however the administrator had not been doing this correctly. Therefore, although the respondent had some logs of the material that had been deleted, the logs could not be said to be comprehensive. These findings are made on the basis of Ms Battersby’s oral evidence which was not meaningfully undermined by anything else in the case. In that scenario, the fact that the respondent does not have a log of the July application (and accompanying emails) being deleted does not clearly show that it was not. Also, it is not in dispute that the respondent was deleting material at the time. In circumstances where we accept that the July application was made but the respondent has no awareness or records of it, the position that is more likely than not is that where the July application and administrator’s email to the client is that they were deleted. This material was all held on the respondent’s old recruitment system. The relevant account was not maintained. We accept Ms Battersby’s evidence, corroborated by chatlogs with the old recruitment system provider, that once the account was no longer maintained it was not possible to recover historic information. There was also no cogent evidence that the material would have been held anywhere else. We accept Ms Battersby’s evidence, in particular, that the respondent made extensive efforts to locate the missing July application and emails and was unable to find them. We accept her evidence because there is no good reason not to.[75]It is right that Ms Battersby merged the two profiles that the old recruitment system had generated for the claimant. This was on 7 August 2023. We accept her evidence that the merger worked as follows. The system would flag two potentially identical applicants based on various data points. It did not rely heavily on the email address used. The primary data point was the content of the CV used. On the basis of the information known, the system merged the July 2022 and April 2023 profiles. Nothing could be safely inferred from the merger about the July application, other than it had not been captured by the system because it did not appear in the merged profile. The merged profile only showed the 2022 and April applications. This finding is made because of Ms Battersby’s evidence, which was not otherwise undermined by anything, and the corresponding records in the bundle of the profile. (v) The August/Michael O’Reilly application[76]The claimant made a further application for the role under a pseudonym namely Michael O’Reilly (‘MOR’) with the respondent’s new recruitment system on 19 August 2023. This application was almost exactly the same as the claimant’s July application, in terms of the CV, other than the use of a different name and references to Nigeria. It was in all material respects the same CV.[77]Although two of the respondent’s witnesses recalled the claimant having also filled in an application form for MOR, we don’t find that this happened. The claimant denied that he filled in such a form. There is no documentary evidence to support its existence. It is the sort of event that, if it had happened, we would have expected documentary evidence to have been generated and retained. There is also no evidential suggestion that anything relation to MOR had been deleted or lost by the respondent.[78]We also accept, however, the respondent’s position that they were unaware of the July application. This is because the respondent has no documentary records of the July application and does not hold a copy of the July CV. Also, the respondent’s lack of awareness is supported by later correspondence by the respondent which does not refer to a July application until the claimant raises it as an issue. In particular, in correspondence between the parties dated 5 December 2023 the respondent carefully set out a comparison of the claimant’s April CV and the MOR CV. It does not mention the claimant having made a July application. If the respondent was aware of the July application then there is no evidenced reason why it would not have also mentioned it in that correspondence.[79]The MOR application was invited for interview after its review, including by Ms Sveinsdottir. The claimant declined to attend that interview.[80]The roles advertised after September 2023 were not filled by anyone else. (vi) The September 2023 application[81]The claimant submitted a further application for the role on 17 September 2023. That application was rejected on 7 November 2023 with the rejection email stating ‘We have ended up moving forward with another applicant’. This was also the disqualification reason marked on the new recruitment platform. At that time the respondent did interview other applicants but no one was ultimately hired for the role. In light of others having been interviewed we do consider the reason given to be necessarily inconsistent. This is because moving forward with another applicant does not have to mean that they were hired.[82]The September 2023 CV contained a further rewording of the claimant’s skills and experience. It also had a cover letter, the final paragraph of which was: Thank you for considering my application. I am enthusiastic about the opportunity to further discuss how my skills and experiences align with Trilateral Research's needs. Please feel free to contact me at [Your Phone Number] or [Your Email Address] to arrange a conversation.[83]The claimant accepted by his own admission that the September 2023 application was not a genuine application, rather it was to provide evidence of what he believed was racial bias in the recruitment process.[84]We find that the reason why the September application was not progressed by the respondent was the clear error and omission in the covering letter. Ms Sveinsdottir told us that although she did not review that particular application, if she had that level of error would have meant that it would have been rejected. The other respondent witnesses gave similar evidence. We agree that the role required attention to detail and this was not demonstrated by the cover letter. Data Subject Access Requests (‘DSAR’s)[85]In addition to the above, on 7 August 2023, the respondent was proposing to delete old applications as part of its transfer between recruitment systems. Applicants were emailed, including the claimant, asking if they wanted to have their data preserved. In response, the claimant asked for his data to be preserved and made a data subject access request (‘DSAR’) on 7 August 2023. The respondent initially responded to the DSAR on 9 August 2023 stating that the respondent held no personal data about the claimant other than the data that had been submitted as part of his application (unspecified) and correspondence between him and the recruitment team. It also stated that he was not shortlisted due to his skillset not matching the respondent’s current needs. The claimant made further DSAR-type queries by email dated 9 August 2023. The parties corresponded about the DSAR for some time, and the claimant was unhappy with the response, and further DSARs were made. These were ultimately responded to by the respondent, in particular, on 19 January 2024 after the claimant made a detailed request for specific documents/information on 23 December 2023.[86]In inter-party correspondence, the respondent by letter dated 5 December 2023 sought to explain the outcome of the various claimant’s applications. This refers to the April and September 2023 applications and not any July application. This is consistent with the respondent not being aware of the July application at that time, it not having been specifically raised by the claimant in correspondence until 15 December 2023.[87]The respondent’s response to the DSAR dated 19 January 2024 included that they had no record of an application made in the name of the claimant in July 2023.[88]We accept the respondent witnesses evidence that the DSARs were responded to in good faith. There is no good reason to find otherwise. Also, we accept that the July 2023 application and email did not form part of the respondent’s response to the DSAR and preservation request because they were unaware of them. We are also satisfied that the relevant members of the respondent did preserve the claimant’s personal information of which they were aware when preservation was requested by the claimant on 7 August 2023. The preserved information is in the bundle. There is no good evidential foundation for the suggestion that it did not preserve information about the claimant that it was aware of. After the claimant’s known personal data had been preserved the company’s main account with the old recruitment system was deleted and the information could not be retrieved after that point.[89]We did not find that anything about the fact of the claimant’s DSAR request and request for preservation of data demonstrated anything relevant to the claims. We do not consider, taking the correspondence as a whole, and when particular information was disclosed to the claimant, that it demonstrates selective disclosure or non-disclosure given our findings on what material was available to the respondent to disclose to the claimant.[90]The respondent has a diverse workforce. In particular, 25% of the sponsorship paid to sponsor applicants from abroad was for applicants who were from Nigeria. This was confirmed in a letter from the respondent’s CEO dated 13 December 2023 and we had no reason to doubt this evidence. D Conclusions 2.1 Did the respondent do the following things: 2.1.1 Reject the claimant’s job applications -2.1.1.1 April 20232.1.1.24 July 2023 (if made) and2.1.1.3 September 2023?[91]In light of our findings above, we find that the respondent rejected the claimant’s applications in April 2023 and September 2023. We find that the claimant made an application in July 2023 but it was either accidently deleted or otherwise lost and it was never meaningfully reviewed. It was also not specifically rejected. 2.2 Was that less favourable treatment? The Tribunal will decide whether the claimant was treated worse than someone else was treated. There must be no material difference between their circumstances and the claimant’s. If there was nobody in the same circumstances as the claimant, the Tribunal will decide whether they were treated worse than someone else would have been treated – a ‘hypothetical comparator’. The claimant relies on the job-applicant ‘Michael O’Reilly’ as a hypothetical comparator. 2.3 If so, was it because of race?[92]We find that rejecting a job application is, in of itself, less favourable treatment, generally.[93]We also find that the non-progression of the July application was less favourable treatment. It was disadvantageous to the claimant because that application was never meaningfully reviewed and therefore was not considered for interview.[94]However, we do not find that the claimant has proved any primary facts from which we could infer that the less favourable treatment was because of his race. This is because there is nothing from which you could properly infer that the claimant’s race played any part in either the rejections or what happened to the July application.[95]The only real argument the claimant had, to suggest that race may have played a part, was the success of his comparator application under the name MOR with references to Nigeria removed. However, in light of the July application never having been meaningfully reviewed because it was either lost or deleted, the comparator is not in the same material circumstances. There is no reason to believe or find that the July application was lost because of the claimant’s race.[96]There is also no proper comparison between the April and July/August CV content for reasons we have outlined above on the differences in presentation and emphasis on skills. Also, between April, August and September the target clients for the respondent changed and therefore there is no meaningful comparison between the three time periods.[97]There is also no meaningful comparison between the April or August MOR application and the September application because the cover letter for September contained significant errors. The success of the MOR application does not indicate anything about the failure of the April, July or September applications given the differences between them.[98]The recruitment exercise was taken in groups of two, and this would mitigate against any individual racial bias. Also, those people had undertaken unconscious bias training. Also, it is relevant that the respondent had a diverse workforce and one of the successful applicants to the April-phase of the exercise was black, Asian or minority ethnic (adopting the language used by the respondent).[99]Even if we are wrong about this, we find that the applications were not successful for the good, proper and cogent reasons given by the respondent’s witnesses. There were non-discriminatory reasons for all of those applications to not go forward for interview. We are satisfied that that the conduct was in no way whatsoever because of the claimant’s race.[100]In particular, the April application was one which presented an applicant with experience of both GDPR and other compliance areas associated with financial services such as KYC and AML. There was more of an emphasis of this range of skills in that application. However, we accepted the respondent’s evidence, in particular of Ms Finn, that at that time they were looking for a pure GDPR-type applicant whose experience was not diluted by other areas. Also, Ms Sveinsdottir’s evidence that she was looking for EU-type experience which the claimant did not have is also relevant. We also find that the September application was tainted by a poor cover letter which contained key errors or omissions.[101]Given that the July application was not progressed by reason of user or system error, there is no reason to believe that any of that was because of the claimant’s race. Rather, we find that it was caused by the respondent moving from one recruitment system to another.[102]Also, there is no factual basis for suggesting that any of the actions of the administrator were because of the claimant’s race, in whole or in part.[103]We consider that the claimant’s focus on purely whether, on the face of the content of his CV, he met the essential criteria and therefore he must have been eligible to be shortlisted for interview suggests that he has taken a very narrow, and mistaken, approach to the respondent’s recruitment processes. The reality was that in deciding whether or not to shortlist an applicant for interview, at any stage, there was also an element of evaluation as to the applicant’s strength and also this was naturally by way of a comparison with others. It was not the case, as a matter of fact, that all applicants who arguably met the essential criteria on paper were shortlisted for interview. In those circumstances, the claimant’s argument that he did meet the essential criteria, and so any subsequent non-progression to the interview shortlist must be for some other unacceptable reason, is flawed. That is not a proper inference that can be drawn from the facts as we have found them to be.[104]We also do not consider that, in reality, the respondent has given inconsistent explanations for his non-progression. Little can be inferred from the recorded reasons for disqualification of applicants because the evidence we heard suggested that these were sometimes done in bulk, were generic, and only a single reason could be used on the system (save for any notes). Given the nature of the role and the scale of the recruitment exercise we see nothing suspicious in this. Ultimately, given that the exact reviewer for each application (other than the August MOR application) was not identified, the later explanations for the claimant’s non-progression could only be given in hindsight. Any argued inconsistency in them does not, in the particular circumstances of this case, give rise to an inference that the nonprogression was in any way because of the claimant’s race. We also accept the respondent’s submission that, in reality, the additional information provided by the respondent about how it has analysed the CVs, and the differences between them as reasons why they are not fair and true comparator examples, has added to the reasons. This is distinct from giving wholly incompatible and inconsistent reasons from which we could fairly infer that the decisions were affected by the claimant’s race.[105]We do not agree, as the claimant submitted, that as a matter of fact there were shifting explanations by the respondent such that we could infer that any treatment was because of race.[106]We do not consider that there was sufficient evidence such that a strength of numbers-type argument would mean that we could infer that any of the treatment was because of the claimant’s race.[107]In the alternative, by the claimant’s own admission, the September 2023 application was not a genuine application. Rather, it was conduct designed to generate evidence in support of the claimant’s claims. In those circumstances, we do not consider that this amounted to less favourable treatment. This is because it was not a genuine job application, as opposed to the April and July applications. The facts of this case are slightly different to Keane, above. We accept that the claimant was genuinely interested in the position when he made the earlier applications. However, in light of his evidence on this topic, we find that he did not suffer a disadvantage when the September 2024 application was not progressed. This is because the purpose of that application was not, in reality, to secure a job, but to generate evidence that the claimant could later rely on. Notwithstanding the outcome of this case, the claimant achieved his objectives with the September application. It is not a disadvantage in those circumstances.[108]For all of the above reasons, the claim is dismissed.