THE AI EMPLOYMENT LAW REGISTRY
Skip to registry entries
The AI Law ReportsLegal reports & articles by Matthew Lee
Subscribe to my work
Ad / marketing communication. Legal reports and articles connected with my practice as a barrister. General information, not legal advice. Publication notice.
Work · Recruitment · Workplace disputes
AI Employment Law Registry
My registry of legal developments concerning artificial intelligence and employment law. Explore the cases, read the linked sources and follow my analysis.
The registry
Cases & developments
Browse the published entries below. The table preserves the spreadsheet’s own columns, with source and commentary links where recorded.
52 published entries
Saved copy · 6 November 2026
On smaller screens, scroll sideways within the table to see every column.
| No | Date | Case | Key Takeaway | Analysis |
|---|---|---|---|---|
| 52 | 18/08/2026 | Slack v Cineworld Cinemas Ltd: 1603220/2025 | AI-assisted witness evidence. In determining whether the claimant was disabled under the Equality Act 2010, the claimant explained that he had written out his “life story” and asked AI to condense it into a witness statement. The Tribunal did not know what instructions had been given to the AI or the extent to which it had constructed or modified the claimant’s words. It therefore considered the statement but treated it with caution, using the claimant’s oral evidence and cross-examination to supplement and cross-check it. The Tribunal ultimately found that the claimant had not established that he was disabled; the AI issue was evidential rather than determinative. | |
| 51 | 14/08/2026 | Katkoria v Stena Drilling PTE Ltd | AI used in contemporaneous employment evidence. In a race discrimination claim concerning redundancy and alternative employment, the Tribunal recorded that a comparator had used AI, together with his memory, to identify examination dates when providing the employer with details of his qualification history. The comparator’s qualifications and subsequent appointment were relevant to the claimant’s case. The Tribunal dismissed the discrimination claims and made no finding that the AI-derived information was inaccurate or that the use of AI influenced the selection decision. | |
| 50 | 10/08/2026 | Buchana v CA | Concerns about an employer’s AI pilot, which recorded client calls and used AI to convert them into written case notes, gave rise to protected whistleblowing disclosures about data protection and client confidentiality. The claimant’s whistleblowing detriment and automatic unfair dismissal claims were nevertheless dismissed because the Tribunal found the protected disclosures did not materially cause the alleged detriments and were not the sole or principal reason for dismissal. | |
| 49 | 05/08/2026 | Williams-St Louis v West Berkshire District Council: 6015012/2024 | AI-generated transcripts and admissibility. The claimant sought to rely on transcripts of covertly recorded workplace conversations which had been produced using AI. The Tribunal refused permission to rely on them where they had been disclosed late and the respondent had insufficient opportunity to check the transcripts against the underlying recordings. Admission would have required postponement of the final hearing. The decision was based on case management and the overriding objective, rather than any general rule against AI-generated transcripts. | |
| 48 | 21/07/2026 | T Dudhill v Bakkavor Foods Ltd: 6046950/2025 | The Employment Tribunal referred expressly to the June 2026 Presidential Guidance recording a significant increase in interim-relief applications, often indicating AI use, accompanied by much greater volumes of documentation. The Tribunal held that the claimant had presented the complaint when he lodged his application to amend and therefore had jurisdiction to entertain the interim-relief application. There was no finding that this claimant personally used AI. | |
| 47 | 01/07/2026 | M Mostafa v (1) K Julier (2) C Slater [2026] EAT 122 | Challenging an Employment Tribunal judge sitting without lay members, the appellant relied on an authority which neither EAT judge could locate. Judge Keith found that it did not exist and queried whether AI had generated it. AI provenance was not established. The renewed application for permission to appeal was refused. | |
| 46 | 18/06/2026 | Opoku v FAB Maintenance LTD | The claimants submitted a 13-page application for reconsideration which the Tribunal said had “the appearance of being prepared with the assistance of AI” | |
| 45 | 18/06/2026 | Opoku and Mr E Asare v FAB Maintenance Ltd and EPOS Direct Ltd: 3309074/2024 | A 13-page reconsideration application “has the appearance of being prepared with the assistance of AI”. The Tribunal nevertheless considered its substance and refused reconsideration because it amounted essentially to disagreement with factual and evaluative findings. | |
| 44 | 11/06/2026 | A-I Nafornita v HSBC UK Bank plc (6027758/2025) | Admitted reliance on AI research. The claimant had used AI sources for advice, but the Tribunal found fundamental defects in her pleading of race discrimination. That claim and her breach of contract claim were struck out; permission to add indirect race and disability discrimination claims was refused. The decision concerned pleading, merits and jurisdiction, not a sanction for using AI | |
| 43 | 08/06/2026 | CX v Secretary of State for Justice (3303470/2022), reconsideration | (3303470/2022), reconsideration Suspected AI-drafted reconsideration submissions. The Tribunal said the claimant’s long, difficult and highly repetitive documents appeared to have been written using AI, preferring one document in her own words. Reconsideration of the remedy judgment was refused for lack of any reasonable prospect of variation or revocation, not as a sanction for AI use. | |
| 42 | 26/05/2026 | Dr L Midgley v Cwm Taf Morgannwg University Health Board: 1603761/2024 | Disability discrimination proceedings included a complaint concerning the claimant not being permitted to use AI applications including ChatGPT and Microsoft Copilot. The AI-related s 15 Equality Act 2010 complaint was withdrawn and dismissed. | |
| 41 | 19/05/2026 | Okopochini v London and Quadrant Housing Trust: 2413147/2023 and Others | The respondent argued that the claimant had over-relied on AI in formulating his claims and preparing for the hearing. The Tribunal expressly drew no negative inference from this, observing that AI can be a useful tool and noting that the claimant accepted responsibility for the way his case had been prepared and presented. | |
| 40 | 19/05/2026 | Brookes v Executive Telecommunications (UK) Ltd | The Tribunal was warned to exercise caution over the claimant’s authorities because AI appeared to have been used. Authorities did not necessarily support the propositions advanced and two could not be located, causing their authenticity to be doubted. The claimant did not dispute the point. | |
| 39 | 18/05/2026 | John v Nuclear Waste Services Ltd: 3309679/2023 and 3300250/2025 | The claimant’s submissions were particularly lengthy and contained extensive case law, giving the Tribunal the impression that they “perhaps… may have been generated by AI”. The Tribunal nevertheless read and considered them. | |
| 38 | 27/04/2026 | Ms D Abraham v Hound Technology Ltd: 6003691/2026 (Reconsideration) | The claimant accepted using AI to assist with documents, including in the context of a reconsideration application submitted within minutes of an oral decision. The Tribunal made no finding of improper AI use but said a primary judicial concern is that AI-generated documents should “reflect rather than suggest a party’s case and evidence”. | |
| 37 | 24/04/2026 | Mr A Smith v ROC Group Ltd: 6008824/2026 | In an interim-relief application, the claimant provided a 192-page bundle containing considerable repetition and accepted that he had received assistance from artificial intelligence in preparing his material. The Tribunal recorded the AI use but made no finding that using AI was itself improper. | |
| 36 | 24/04/2026 | Thompson v Quiver Delivery Ltd: 6021161/2025 | In an indirect race discrimination claim concerning a requirement for job applicants to submit an introductory video, the claimant relied on a screenshot of a Google search which provided “nothing more than an AI summary” of unrelated material. The Tribunal found there was no concrete evidence establishing the alleged group disadvantage and struck out the claim. | |
| 35 | 21/04/2026 | Kolaniak v Nice Systems UK Ltd: 6007192/2025 | AI featured directly in a redundancy and unfair dismissal case. Following the claimant’s dismissal, her responsibilities were absorbed by other team members and the Tribunal accepted evidence that the employer had, with the help of AI, successfully managed the workload without replacing her. The dismissal was nevertheless unfair because the decision was predetermined and consultation was ineffective. | |
| 34 | 20/04/2026 | Smith v Addingtons Associates Limited: 3314144/2023 | An AI-generated transcript relevant to an issue in the proceedings was found to make no sense in parts. The Tribunal ordered the parties, by consent, to provide an agreed transcript instead. | |
| 33 | 27/03/2026 | Bull v Smarta Energy Ltd: 6002192/2025 | Employment Tribunal rejected an argument that a claimant’s use of AI to help prepare a withdrawal meant that he lacked agency in making it. The Tribunal said it was his right to use AI to prepare submissions, but he remained responsible for the prompts, information and resulting document. | |
| 32 | 23/03/2026 | Saatci v Richmond Food & Supply Ltd | The Employment Judge used Microsoft Copilot to translate parts of a reconsideration application submitted in Turkish. The Judge also considered it very likely that AI software had been used in preparing the application. | |
| 31 | 26 January 2026 | Sunder v Royal Mail Ltd and others: 1401270/2024 and 2226866/2024 | Suspected or Confirmed Hallucination. See Tracker | |
| 30 | 20/01/2026 | Mr R Venugopal v One Stop Stores Ltd | Claimant said he could not afford lawyers and instead used AI for legal research. This seems to have led to a lack of clarity | |
| 29 | 09/01/2026 | Mr R Roberts v RELX Group plc and Others | An issue in the case was a recording (or rather the use of AI to produce a transcript) being made by an individual working in the Philippines. | |
| 28 | 07/01/2026 | Mr G Severi v Bridewell Consulting Ltd | AI-generated meeting transcripts were in evidence, but the tribunal said it was inaccurate in places and treated it cautiously | |
| 27 | 30/12/2025 | Murphy v SOS for Work and Pensions: 2400194/2024 | With the assistance of AI, Claimant produced numerous documents purporting to clarify and expand his claims, which contributed to confusion. The judge warned against “firing off AI generated documents and applications” in a manner inconsistent with the overriding objective. | |
| 26 | 19/12/2025 | Christopher v Maximus UK Services Ltd | The judge said one further-information document looked AI-generated and the claimant accepted earlier attempts had been AI-assisted. Tribunal stated that if it had not beeen used for a certain document then it was certainly elongated by the setting out at the beginning and the end of totally unnecessary quotations. | |
| 25 | 11/12/2025 | Sullivan v Capita plc: 6039320/2025 and 6035838/2025 | Claimant had used AI software to construct the main wording of withdrawal message, and did not fully understand the legal implications of the withdraw | |
| 24 | 08/12/2025 | Iwobi v City of London Academies Trust | AI used to draft witness statement | Link |
| 23 | 02/12/2025 | O Harwood-Allen v London United Busways Ltd | Suspected or Confirmed Hallucination. See Tracker | |
| 22 | 24/11/2025 | Ms A De Vere v W H Malcolm Ltd | It was put in cross examination that ChatGPT had been used to create documents as they had hallmarks of AI. This was denied and tribunal took individual at their word. However, the documents did nothing to clarify the claims. | |
| 21 | 21/11/2025 | Caird v British Telecommunications plc | An appeal hearing was recorded by laptop which generated a transcript. That transcript was not before the Tribunal, but rather a document headed ‘Appeal Meeting Summary Notes’ that had been generated from the transcript by the use of Microsoft Co-Pilot. | |
| 20 | 18/11/2025 | Wlochinski v English Lakes Hotels Limited | Tribunal suspected lengthy submissions were partly AI-generated and poorly edited. Noted it is not proportionate for the Tribunal to wade through such material. | Link |
| 19 | 14/11/2025 | Hunter v SmartFrame Technologies | Claimant accepted using artificial intelligence to assist (editing) the chronology and particulars of claim. Tribunal also noted difficulties with automatically generated transcripts. | |
| 18 | 11/11/2025 | Heath v NHS England | Claimant used AI (ChatGPT/Grok) to generate an overlong list of issues. Tribunal said no issue in principle, but it made focus and justification harder. | |
| 17 | 20/10/2025 | Cooney v WPP 2005 Ltd | The Claimant said document represented the product of an instruction the Claimant gave to an AI tool to provide information on that subject, and the Claimant noted that the sources that were examined by the AI programme to produce that document were listed at the bottom of it. | |
| 16 | 09/10/2025 | Peters v DVSA | Case-law ‘summaries’ were generated with an AI tool and had significant inaccuracies. The Tribunal could not rely on them and this caused or contributed to the judgment being reserved. See AI Hallucinations Tracker | |
| 15 | 30/09/2025 | Whelan v GenSight Ltd | Claimant acknowledged using AI to draft a 103‑page witness statement. Tribunal criticised the result as unwieldy and urged concise, issue-focused evidence in the claimant’s own words. | |
| 14 | 17/09/2025 | Tsikni v Kontis and Alphakon Ltd | Complaints about use of ChatGPT/AI-generated content. The use of AI was taken into account on allocating weight to the evidence | |
| 13 | 08/09/2025 | Murly-Cleves v University Hospitals Sussex NHS Foundation Trust | ||
| 12 | 15/08/2025 | Mrs Anna Kuzniar v General Dental Council | Claimant said she used ChatGPT/AI to create submissions and citations. Many were incorrect or non-existent. See AI hallucination Tracker. | |
| 11 | 28/07/2025 | Chanda v Royal Mail Group Ltd | False/mischaracterised case citations. Tribunal thought they may have been AI-generated (though denied). See hallucination tracker (ADD TO MAIN) | |
| 10 | 02/06/2025 | Walker v Polygon Public House Limited | Claimant sent a document he said was AI-generated (checked by a lawyer friend) giving ‘hearing advice’ and suggested questions that may be asked in cross examination. Tribunal said it did not assist. | |
| 9 | 11/04/2025 | Akbar v FPKglasgow Ltd | Claimant said she used ChatGPT to research her legal rights and to start the claim as a litigant in person. | |
| 8 | 14/02/2025 | Nicholas v Rabart Decorators Merchants Limited | Use of AI tools as auxiliary aids. Grammarly Pro and ChatGPT. | |
| 7 | 19/12/2024 | Wright v SFE Chetwode Limited and Winter | Claimant used ChatGPT to write her statement and submissions (adding her own details). Tribunal declined to debate AI use but stressed she must personally understand the case. | |
| 6 | 09/10/2024 | Messi v Change Grow Live and Others | Evidence use. Documents included AI-generated transcripts of recorded phone calls made using a mobile app. | |
| 5 | 01/08/2024 | Rollo v Marstons Trading Ltd | Claimant sought to rely on ‘ChatGPT statements’. Tribunal treated them as unhelpful (not expert evidence) in assessing the application. | link |
| 4 | 16/02/2024 | Lwanga v Atkins Limited | Claimant used ChatGPT to produce written material for the case. Tribunal noted the limitations of generic AI output when assessing relevance and presentation of submissions. | |
| 3 | 28/11/2023 | Warrener-Iglesias v Coretech Security Services LTD | Claimant used ChatGPT to help structure his witness statement. Tribunal assessed evidence on its merits rather than treating AI assistance as disqualifying, but accepted that the fact is was not in Claimant’s own words means it should be given less weight. | |
| 2 | 16/10/2023 | Manjang v Uber Eats UK Ltd | Algorithmic/automated decision-making was central. Reconsideration concerned whether deactivation was by a solely automated decision and related factual/procedural issues. (ADD to eq and bias) | |
| 1 | 09/07/2022 | Manjang v Uber Eats UK Ltd and Others (Preliminary Judgment) | Algorithmic/automated decision-making in a platform context. Issues included how the platform’s processes operated and their legal implications for the claims. |
No entries match your search.
Read each entry in context. Inclusion does not establish AI use, wrongdoing or liability. Check the underlying material and any later developments before relying on an entry. Dates and summaries are reproduced from the registry. Publication notice.
How the registry updates
This table reads the published Google Sheet when you visit and checks for changes every minute while the page is open and visible. It also checks when you return to this tab. “Refresh entries” checks immediately. Google may take a few minutes to republish edits. Case and analysis links are retained.
If the live source cannot be loaded, the page identifies the saved copy being shown. Blank cells are preserved from the source spreadsheet.
About this research
Research by Matthew Lee
I maintain this registry as part of my research and writing on artificial intelligence and the law. It forms part of my legal reports and articles, alongside my wider analysis at Natural & Artificial Law.
The registry is a starting point for exploring the sources and their legal context. It is a selected research collection and is not exhaustive.
Follow the developing law.
My legal reports, articles and updates from Natural & Artificial Law.
Publication information
Ad / marketing communication. General information, not legal advice. AI tools may assist this publication; errors, omissions and fabricated references can occur. Check the underlying primary sources. No Public (Direct) Access. All instructions via my clerks at Doughty Street Chambers.
Read the full publication notice
Ad / Marketing communication. Not legal advice.
The AI Law Reports and Natural & Artificial Intelligence in Law are my professional legal publications. Their reports, articles, registries and trackers form part of my practice marketing and support my teaching, lecturing and writing. They provide general information and commentary, not legal advice in any jurisdiction.
Reading, subscribing or making contact does not create a retainer or barrister–client relationship. No duty of care is assumed in respect of these publications.
AI use, accuracy and checking sources.
AI tools may assist research, drafting, checking, formatting and presentation. Human review cannot guarantee that every error has been identified. Inaccuracies, omissions, outdated material and fabricated references may remain.
Verify cases, quotations, citations, dates, legal propositions and any later developments against the underlying judgments and other primary materials. Completeness, accuracy and currency are not guaranteed.
Research starting points, not findings.
These are not exhaustive or official law-reporting databases. Entries may concern allegations, suspected incidents or unresolved proceedings. Inclusion does not establish AI use, wrongdoing, causation or liability. A summary is not a substitute for reading the full source, its context and the procedural outcome.
Instructions and corrections.
I am a barrister regulated in England & Wales and am not authorised for Public (Direct) Access. All instructions must go through my clerks at Doughty Street Chambers. Do not send confidential or time-sensitive information through this website.
For publication errors, request a correction, identifying the page and passage.