AI: Artificial intelligence in the workplace – are you compliant? Checklist/Guide

Reasons to choose Wilson Browne

Artificial intelligence (AI) has recently exploded in popularity and use. This is largely thanks to chatbots like the ones provided by OpenAI (Chat GPT and GPT4), Microsoft, and Google.

However, the use of AI in the workplace extends well beyond using chatbots – e.g. helping prepare replies to emails, writing articles, or for research. As a result, AI in the workplace interacts with a number of legal obligations in relation to, for example, contractual rights, data protection, and discrimination.

The relatively new emergence and use of AI brings with it a number of “traps” for the unwary which could turn something which was intended to improve efficiency and save costs into a big headache.

The following checklist will help you self-audit and identify whether your organisation needs to investigate its use of AI, and if there is a risk of non-compliance. You can also download a copy here

Does the organisation…	where applicable	If ‘Yes’what are the implications
Use AI to screen CVs (e.g., using an algorithm to sort CVs based on key phrases or criteria)?	Y/N	AI can inherit discriminatory biases. Therefore, if the AI inadvertently discriminates on the basis of a protected characteristic (e.g., sex), a job applicant could issue a claim for discrimination.
Use AI to evaluate performance on job assessments and psychometric tests (e.g., software which uses an algorithm to tailor assessments or score candidates)?	Y/N	In addition to considering the risk of bias which could lead to discrimination claims, employers will need to consider whether their existing privacy notices (if any!) sufficiently cover their use of AI.
Use AI to schedule shifts (e.g., using an algorithm to schedule shifts according to forecasted footfall)?	Y/N	Employees may be given irregular working patterns at short notice. This could entitle an employee to bring a claim of constructive dismissal. It could also impact an employees protected characteristics thus risk claims of discrimination.
Use AI to evaluate and appraise performance?	Y/N	Job applicants have a right not to be subject to automated decision-making under the UK’s GDPR. Additionally, employers must ensure that they are transparent about how performance data will be used in any privacy notice. Employers are unlikely to be able to justify dismissals or performance-related sanctions in the employment tribunal if the decision was made by, or with heavy reliance on, AI.
Use AI to monitor productivity?	Y/N	In addition to privacy concerns, employers must ensure that AI surveillance does not breach the obligation of confidence or contribute to increased stress and anxiety. Additional considerations will be required when dealing with performance management or sickness absence processes/procedures.
Use or allow the use of AI chatbots (generative AI, e.g. ChatGPT)?	Y/N	Employees using AI chatbots could inadvertently breach confidentiality obligations related to confidentiality or inadvertently inform/advise if the information supplied by the chatbot is not properly vetted. This should be built into employee training, disciplinary and performance management processes/procedures. Employers may also wish to ensure that any materials generated by or with the help of AI can be copyrighted, which may not be possible in every instance.

Any “yes” answers suggest that a proper investigation into the use of AI may be necessary.

All liability for the use of AI in a way which is not legally compliant sits with the employer, irrespective of the provider of the AI software.

In addition, employers may find that they are unable to obtain justification for decisions made by AI while AI software companies remain fiercely protective of trade secrets related to their software. This will make it difficult for employers to defend against claims related to the use of AI in the workplace.

How can the Wilson Browne employment team help you?

We advise employers of all sizes on the full range of employment law issues and, as a small, closely knit team, we are proud of the personal relationships we build with our clients.

Our employment team can assist employers develop strategies for the fair and lawful use of AI in the workplace, implement policies to address the use of AI, and advise on any employee relations issues arising from the use of AI.