Automating the board? An English Law Perspective
There has been a great deal of publicity relating to algorithms or artificial intelligence being “appointed” to company boards, but what roles can AI have in the boardroom, given the current legal framework and technology, and how can companies make use of AI within their corporate governance arrangements?
In a recent blog post we considered this question from a South African law perspective. Below we give an English law view.
While it might make an attention-grabbing headline, it isn’t possible to formally appoint AI software as a director of a company incorporated in England and Wales.
Under section 155 of the Companies Act 2006 (2006 Act) a company may have a legal person as a board member provided it has at least one director who is a human. Under English law, AI is not a legal person. There have been various outline proposals to grant AI legal personhood, but at present (and for the foreseeable future) English law would view AI as no more than a tool, like any other software.
In addition, a proposed amendment to the 2006 Act (section 156A), not yet in force, would require that directors of a board are human (with few exceptions).
Directors may of course delegate their decision-making powers, so could they delegate to AI?
The 2006 Act codifies various common law duties of directors. Both the duty to exercise independent judgment in section 173 and the duty to use reasonable care, skill and diligence in section 174 could potentially be relevant if a director delegated decision-making authority to AI without coming to their own view.
Similarly, in the case of banks and financial institutions, outsourcing regulations in many jurisdictions provide expressly that a bank or financial institution cannot transfer responsibility for regulatory compliance by outsourcing it. The Senior Managers Regime in the UK makes this clear in the case of their senior managers too.
All this points to a clear dividing line between when it is acceptable to use AI as a tool in decision-making, on the one hand, and when the use of AI constitutes an abdication of responsibility, on the other.
But this is not to say that AI has no place in the boardroom. Data is often held in silos in businesses and AI along with data analytics can mine that data to deliver new insights for directors and managers.
AI will result in a move away from time-consuming administration for senior management and will also be increasingly used in checking for, and ensuring, regulatory compliance in relation to proposed business decisions.
At least in the short term, therefore, the role of AI at board level will be to augment, test and verify human decision-making rather than to replace it.
Perhaps most importantly, AI will help management to concentrate on value creation through factors that are inherently human. Management will be freed up to spend more time on activities involving human judgment, people coaching and development, and tasks requiring emotional intelligence and creativity.