The EAT’s decision in Hancox v Sutherland and Others is one of the first Employment Appeal Tribunal cases to deal directly with a litigant in person’s use of generative AI.  While the appeal itself concerned a strike out decision, the wider significance of the judgment lies in the guidance it provides on the responsible use of AI in employment litigation, which is also relevant to employees using AI for grievances at work. Employees representing themselves in Employment Tribunal (& at work) are increasingly being seen to use generative AI tools, such as ChatGPT to draft submissions, grievances, and get legal advice. This is resulting in huge documents which are often inaccurate or misleading.

In this case, Mr Hancox brought various claims against four respondents.

Following the dismissal of an interim relief application, he embarked on a LinkedIn campaign criticising the respondents and publishing information relating to the tribunal proceedings, including disclosed documents. The Employment Tribunal subsequently struck out his claims, finding that his conduct had been scandalous, unreasonable and vexatious, and that a fair trial was no longer possible.

On appeal, Mr Hancox submitted a 300-page skeleton argument generated using ChatGPT. The document did not comply with the EAT’s Practice Direction and, at the hearing, Mr Hancox accepted that he had not checked it properly for accuracy and that it lacked credibility. He had also asked the Respondents to check it for him.

Although the EAT allowed one ground of appeal to proceed to a full hearing, the most important aspect of the judgment was the guidance issued regarding the use of AI-generated documents. The EAT confirmed that litigants who use AI should ensure that any documents submitted to a tribunal comply with applicable procedural rules, are checked as thoroughly as reasonably possible for accuracy, including factual assertions, evidence, legal arguments and focus on relevant issues and avoid unnecessary repetition.

Importantly, the EAT emphasised that while litigants in person are not subject to the same professional and regulatory obligations as solicitors and barristers, they are still personally responsible for the content of documents submitted to a court or tribunal.

Employment lawyers have become increasingly familiar with receiving AI-generated pleadings, witness statements and applications from litigants in person.  Similarly, employers are seeing many more grievances from employees assisting by or generated by AI. Such documents can often be lengthy, repetitive and, in some cases, contain legal authorities or propositions that are inaccurate or just completely wrong.

The EAT’s judgment makes clear that a litigant cannot escape responsibility by blaming an AI tool. Submitting unverified AI output and expecting either the tribunal or the opposing party to identify and correct errors is not acceptable.

For employers the decision is likely to prove helpful when dealing with AI-generated documents. More broadly, Hancox signals that employment tribunals are willing to embrace the use of AI, but only where it is used responsibly. The message from the EAT is clear, AI can assist with drafting, but the person submitting the document remains responsible for ensuring it is accurate and fit for purpose.

If you are an employer and want to update your grievance policy to deal with handling AI generated grievances, please contact the team.