Artificial Intelligence in the boardroom?

Some think that one day robots will take over the tedious work that still keeps many humans occupied on a day to day basis. But what about having a robot on a board of directors? Isn’t that the stuff of science-fiction?

Apparently not. Deep Knowledge Ventures (DKV), a Hong Kong based venture capital fund which invests in life sciences, cancer research, age-related diseases and regenerative medicine, has appointed an artificial intelligence algorithm called VITAL (Validating Investment Tool for Advancing Life Sciences) to its board of directors. DKV insists that this is more than a publicity stunt: VITAL is capable of providing the board with data analysis that can absorb and evaluate efficiently huge amounts of data, is unbiased, disinterested and has no problems to reason logically. It can also pick up market trends not immediately obvious to humans.
DKV’s board will not make any decisions without taking the advice of VITAL into consideration and the algorithm has been given a seat on the board. The long term goal of DKV and VITAL’s creator, Aging Analytics (a UK company) is that VITAL be able to make completely autonomous decisions and to vote independently on company business.

Legal technicalities aside, one could argue that VITAL is just a piece of software and lacks any qualities that a “proper” director would have. But AI programmes called “learning cognitive agents” develop fast and can already converse with humans, learn, understand the meaning of words and sentences and make suggestions or provide answers to complex questions. So who knows where this could lead for a programme like VITAL?

Where does that leave the law?

UK law allows legal entities other than natural persons to be company directors e.g. bodies corporate. Understandably, there is no provision about intelligent machines which are certainly not recognised as legal entities in their own right.

It is interesting to speculate whether it is possible or desirable for the law to evolve in this area. Let’s make the big assumption that humans could even perceive an advanced cognitive agent like a more personable version of VITAL as a person and not a mere tool. Lots of theoretical and practical questions would arise. For example, in what realistic sense could a machine be held accountable for its decisions in the way that human directors can be? How would a machine bind the company, e.g. conclude contracts or dismiss employees? Can a machine retire, get “ill” or make legally acceptable excuses for poor performance or violation of regulations?

The UK government is proposing to introduce a general ban on corporate directors. One of the reasons for this is to ensure transparency and accountability with respect to the individuals influencing the company (which assumes, of course, that a human director is more transparent and accountable than AI, which is philosophically interesting). So it seems we are currently moving in the opposite direction from the futuristic vision of AI on the board. But who knows?! Watch this space …