Whilst AI has many benefits in the workplace, we are seeing many employers now grappling with lengthy, complex grievances that have clearly been drafted with AI.  These are often much longer grievances and frequently do not make an awful lot of sense, being repetitive or vague, making it difficult for employers to establish the core of the employee’s complaint and understand what the real issues are.  Unfortunately, it is not uncommon, in our experience, for the employee themselves not to have fully reviewed their submission and understand everything that has been included.

Every grievance needs to be properly investigated and responded to by an employer and cannot be discounted purely on the basis that it is difficult to understand.  If such a complex or lengthy AI driven grievance is received, it is recommended to hold an initial ‘scoping meeting’, during which the employer can ask questions to understand the nature of the grievance, in the employee’s own words, and a summary of the grievances can then be produced and agreed before the investigation moves forward.  This helps to ensure that any subsequent investigation, grievance meeting and outcome can focus on the material issues.  This should hopefully assist to flush out where AI has ‘hallucinated’ or embellished any issues.

A scoping meeting should only generally be used on those occasions where the nature of the grievance is not clear (so as to avoid the burden – in terms of time and resources - of holding an additional meeting). However, this may lead to some employees complaining that they are being treated differently to others, and it may not always be clear that the grievance outcome letter reflects the actual complaint unless this has been clearly clarified.

It is therefore recommended that employers amend their grievance procedures to reflect:

• that a scoping meeting might be held, although it will not need to be held on every occasion;
• suggested limits on the length of grievances;
• prohibiting the use of AI, or imposing limits on the way in which AI is used (for example, that it only assists with structuring and grammar, rather than drafts the entire grievance);
• setting out what key information the grievance must contain, in terms of the nature of the complaint, any relevant facts, dates and names of individuals involved and the resolution that is being sought;
• prohibiting the entry of company confidential information, or information which reveals the personal data of anyone other than the complainant, into an open platform AI tool, so as to avoid exposure of the employer to a data breach.

EAT guidance on use of AI in the preparation of tribunal documents

There has also, unsurprisingly, been a significant increase in the number of tribunal claims and related documentation being prepared, often by claimants who are not legally represented.  In Hancox v Sutherland and others 2026 EAT 139, the employment appeal tribunal (EAT) issued guidelines on the use of AI by litigants (particularly litigants in person) when preparing documents for tribunals.

Mr Hancox’s claim against four individual respondents was struck out by the tribunal on the grounds that he had conducted the proceedings in a scandalous, unreasonable and vexatious manner by conducting a social media campaign on LinkedIn, and that a fair trial was no longer possible. Mr Hancox appealed the decision, and this was listed for a preliminary hearing with the EAT.

At the EAT, Mr Hancox submitted a 300-page skeleton argument created using ChatGPT.  The EAT found the document did not comply with the EAT Practice Direction 2024 (PD), served to "obscure rather than to illuminate" the issues, and was not within the PD's recommended five to 15 pages. The claimant had not checked it and accepted that it lacked credibility.  The EAT was also critical of the claimant’s suggestion that the respondent’s lawyers should review and correct any legal errors within the document after disclosing it at short notice prior to the hearing.

Beyond addressing the use of AI, the EAT took the opportunity to remind all litigants, whether represented or not, of their responsibility to comply with the Employment Appeal Tribunal Rules 1993 (as amended) and the PD, to assist and not mislead the tribunal.

The EAT recognised that there is potential benefit to be gained from the effective, responsible and ethical use of AI, including by those who do not otherwise have access to professional legal assistance.  However, the EAT also pointed out the risks, including that the output can be unfocused, excessive or not of direct relevance; does not comply with procedural requirements; is the product of inadequate/inaccurate input material or poor prompting; is misleading or does not reflect the true factual, evidential or legal position in a case; raises new points on appeal that were not raised at first instance; is affected by in-built biases; or contains hallucinations (for example made-up case citations).

The EAT stated that litigants who use AI must ensure, at a minimum, that all documents submitted:

1. Comply with applicable procedural rules;
2. Have been checked as thoroughly as the litigant or representative is reasonably able for accuracy, ensuring that factual, evidential and legal points (including references to authorities) are correct and that the court or tribunal is not misled; and
3. Contain only relevant points, with the focus being on the central or best arguments, presented in an easily comprehensible manner and avoiding undue repetition.

They went on to say that simply submitting the product of AI to a court or tribunal or placing the onus to check a document onto an opponent, is not acceptable.  Where concerns arise that necessary checks have not been undertaken, or that they have been conducted inadequately, judicial enquiries and potential sanctions are likely to ensue.

The case is certainly not the end of complex and lengthy pleadings and submissions being made by claimants in person, but it is hoped that this is a useful case to ensure that appropriate sanctions can be made where AI is used in a misleading or overly complex manner.

 For further information or to discuss how you may need to amend your grievance procedures, please get in touch with KKnox@ortolan.com or jjones@ortolan.com