AI employees: is the world of work ready?

Last month, an American HR software company assigned three AI-powered ‘digital workers’ with employee records, training, performance targets and a human line manager. The scheme sparked much debate and was withdrawn after only three days, raising important questions for all employers – what is an ‘AI employee’ and is the world of work ready?
From a legal perspective, the current position is that AI tools (as software or data) cannot hold the legal status of ’employee’ in the same way that human employees can. AI tools carrying out the work of employees cannot therefore enter into an employment contract, or qualify for employment rights such as the right not to be unfairly dismissed, or the right not to be subjected to unlawful discrimination.
A lack of legal status doesn’t prevent employers however from being creative and labelling and treating its AI tools as ’employees’ in name only, inviting practical, legal and potentially ethical considerations for all employers.
Take for example a customer service chatbot: from a practical perspective this could feasibly sit within a customer service team, have a place on the org chart, a human line manager and be assigned performance targets for number of calls answered, queries resolved, tone and language used.
An AI employee’s capabilities would of course be vastly different to its human counterpart, remembering that AI tools don’t need to take legal rest breaks, annual leave or even sleep. Any performance concerns could be a simple issue of reprogramming, rather than any performance management process.
Given that some AI tools have the potential to exhibit or learn bias which can result in (unintended) discriminatory outcomes in their decision making processes, it will be important for employers to understand any such potential risks if embedding AI employees within its workforce. Particularly as liability for any discrimination claims may fall with the employer and not with any AI employee.
Another important consideration is the potential for impact on employee morale and the wider perception of AI tools within the workforce.
On the one hand, AI employees could relieve pressure on existing teams and create a positive workplace culture of working together with AI towards common goals. However, at a time when surveys suggest that many employees hold concerns that AI will replace them, introducing AI employees could risk unsettling the workforce, or leave employees feeling as though they are in direct competition with AI.
With the new Labour government committing to putting ‘worker voice at the heart of Britain’s digital transition’ this could potentially give employees more of a say in how AI tools are integrated. Whether this would extend to matters such as AI tools being positioned as AI employees, we don’t yet know.
There is clearly plenty for employers to think through before AI employees enter the workforce and it will be interesting to see if, and how, the concept evolves and what further issues arise along the way. In the much longer term, we may even see AI obtaining legal employee status, be able to join a union and even pursue claims against its employer – but this remains to be seen.