Dealing with employee grievances in an AI-enabled world

by Katherine Pope, Partner at BCLP

Workplace grievances written with the assistance of generative AI are on the rise, increasing the pressure on HR departments and line managers. Katherine Pope offers guidance to help employers deal fairly and efficiently with lengthy, legal citation-rich documents that sometimes obscure rather than reveal the core issues.

Individuals and businesses increasingly turn to AI for advice across many areas of life, and the workplace is no exception. Employees are using AI tools to formulate workplace grievances, with practical implications well known to any HR professional who has received one: the content may be longer and less focused, citing legal provisions rather than clearly setting out the employee’s own concerns. AI-generated grievances can also contain errors and inaccuracies, as AI tools can produce lengthy output from limited prompts and context.

It is important that employers are able to deal with grievances effectively. Ultimately, the aim is to resolve the grievance fairly, spending a proportionate amount of time on the matter.

How can this be achieved?

Tips for employers

AI-generated grievances may set out complaints at length without articulating what the employee actually wants to achieve. If that appears to be the case, remember to ask the employee to identify the specific outcome they are seeking. Understanding the desired resolution early in the process can help focus the investigation and make it easier to reach a meaningful outcome.

When reviewing the grievance document

• Don’t be deterred by the volume or apparent legal sophistication of an AI-generated grievance. References to legislation or legal concepts do not necessarily mean the complaint is legally complex. Strip back the language and focus on the substance: what has happened, and why is the employee unhappy?
• Be alert to inaccuracies. AI tools can generate content that sounds authoritative but is factually or legally wrong, for example citing legislation incorrectly or mischaracterising internal policies. Do not assume legal references are accurate. Where legal points are raised, take advice if needed, but do not feel obliged to engage with every legal assertion in the grievance.
• Consider whether the employee genuinely understands and endorses the full content of their grievance. Where a document is clearly AI-generated, it can be helpful to check (sensitively) that the employee is familiar with everything raised. This is not about discouraging AI use, but about ensuring the process focuses on the employee’s real concerns.

At the hearing and outcome stages of the process

• Don’t lose sight of the fact that the purpose of a grievance process is to resolve the issue complained of. Understanding the core of the complaint is essential. It may be helpful to use the grievance hearing to encourage the employee to set out their concerns in their own words. Explore the main themes and core issues, and seek to agree them with the employee.
• Keep the process and outcome proportionate. An AI tool can generate a lengthy document from a short prompt, and the resulting grievance may contain repetition, tangential points or issues the employee does not feel strongly about. There is no obligation to address every line individually. Instead, group concerns into themes and address each substantively in the outcome letter.

It can be helpful for the employee handbook and grievance policy to make clear that the employer will be best placed to understand and resolve a concern if the employee articulates it clearly, and in their own words.

Separately, consider refreshing IT and systems policies to make clear that confidential business information should not be uploaded to public AI platforms, and to remind employees of the importance of complying with the organisation’s acceptable use policies.

Employer use of AI tools

Employers may themselves use AI tools to summarise and respond to grievances. Those involved must comply with the organisation’s guidelines on AI use, and consider whether the platform is approved for use with personal data.

When using AI tools to draft grievance responses or outcome letters, prompts should be carefully and professionally worded. The prompts themselves (not just the outputs) may form part of the disclosable record in the event of a data subject access request or litigation. Avoid including subjective opinions or commentary about the employee in any prompt.

Wider impact of AI in workplace disputes

Employment tribunals face similar challenges, with reports of increased AI use by unrepresented parties leading to the drafting of overly complex claims, responses and applications. This is creating a new form of pressure on the already stretched tribunal system. Official guidance for judicial office holders notes that material from unrepresented litigants may be generated by AI without verifying the accuracy of its content, and warns of the possibility that deepfake technology may produce fake material including text, images and video.

There are clear parallels with the challenges facing HR professionals. The judicial guidance to be alert to AI-generated content, look beyond the apparent sophistication of drafting, and verify accuracy, applies equally to internal grievance processes.

AI tools are here to stay, and their workplace use will only increase. That is not something to resist, but it does require adaptation. The fundamentals of good grievance handling have not changed: listen to the employee, understand the complaint, investigate fairly and reach a reasoned outcome. By staying focused on those fundamentals and being alert to the risks AI-generated content can introduce, HR professionals can continue to manage grievances effectively and proportionately, however they are presented.