In the last few years AI has become increasingly widespread in UK workplaces. As it becomes more embedded in everyday working practice, it will also become a common feature of Employment Tribunal disputes. Although there has only been a trickle of cases involving AI reaching the Employment Tribunals so far, that's likely to increase (just as employees' use of social media has spawned a whole body of sometimes-controversial caselaw).
Although the UK has yet to introduce specific legislation to regulate AI in the workplace, key employment law concepts such as unfair dismissal and discrimination will be used to scrutinise how it operates. Given concerns about the reliability and transparency of AI tools, as well as their potential for perpetuating bias, this is likely to be fertile ground for employment disputes. Some of the issues we expect the ETs to be grappling with are:
- How much detail employers need to give about AI-enhanced decisions - and how much employers are expected to interrogate the AI tools they buy in to understand how the decisions are made.
- Whether AI tools can be relied on when being used for unintended purposes - e.g. a meeting summary tool designed for recruitment being used in a redundancy process. Will this undermine the fairness of any decisions?
- How far employers are entitled to trust technology which they have decided to use for management purposes - will this be within the “band of reasonable responses” from an unfair dismissal perspective?
- And, of course, the risk of bias and inaccuracy, whether from poor design or poor underlying data. (How many employers are doing a data audit before they buy in solutions which will depend on their existing data?)
Although AI has huge potential benefits for employers, there are also some under-appreciated risks. There's always a temptation for suppliers to over-promise, particularly about tricky issues like privacy and data protection. And when employers feel that they're behind the AI curve, there's a tendency to buy in solutions hastily before they have been properly evaluated.
At the moment, there is a real lack of practical guidance for employers trying to navigate this area. The Government has published guidance on using AI in recruitment, but it's not clear how many employers are using it. Although we don't underestimate the complexities of producing more detailed guidance (or the challenges of staying ahead of technological developments), a Code of Practice, incorporating guidance from the ICO and EHRC, would be hugely useful. In the meantime, employers should be alive to the potential pitfalls as well as the benefits.
Want to hear more from us? 
