Artificial intelligence increasingly permeates our daily lives. What may have started with enlisting help to plan a travel itinerary for a weekend away can evolve into entrusting AI with more serious scenarios such as workplace difficulties.

There is growing evidence that employees are using AI to draft their grievances about staff related issues. This can result in much longer documents, reliance on inapplicable law, possible data breaches and an overall style that is simply not typical of that employee’s writing.

This trend can place undue strain on HR functions which may be ill equipped to process lengthy complex grievances. Larger employers may need to train their HR professionals to be alert to AI generated grievances and address them in a reasonable and concise manner. This largely involves stripping away what might be flowery or irrelevant material and focusing on the core of the complaint to ensure the employer’s response is compliant with good practice and its own policies.

An immediate concern when an employee is using a large and open language model is that the employer’s confidential information and the personal data of individuals might be uploaded to unknown servers and be absorbed by a learning algorithm. This can mean sensitive data becoming public knowledge and possibly even used in the chatbot’s unrelated answer to another question.

The employer’s AI policy should therefore be alive to this and should clearly prohibit uploading confidential information or personal data to open AI tools. For the employee, using AI to take advice on a workplace dispute or to draft a grievance or Tribunal claim may therefore place them in breach of contract entitling the employer to take appropriate action. It could prove entirely counterproductive for that employee.

Another clear danger of a disaffected worker turning to AI for an answer is that the tool used might provide a slanted view on merits particularly if it is only briefed as to one side of the dispute. This can embolden the staff member with a potentially false view of the strength of their case.

Staff should therefore be very careful about using open AI tools for work related issues as it is quite likely to be in breach of an employer policy. Meantime, employers faced with the product of any such AI referral should immediately be alive to the possible breach of confidentiality that has taken place. In some cases the employer’s response will have to be balanced against the statutory protection an employee gains when making a qualifying disclosure, which grievances can often constitute. Over time, case law will likely shape clear guidance as to how employers should react in these circumstances but for now, as AI increasingly wields its influence, each case has to be assessed on its nuanced merits.

John Aycock is head of M&P Legal’s employment team with over 35 years’ experience of labour law in three jurisdictions. Note this article does not constitute legal advice, specific advice should be sought on the facts on each case.

Back to top