The Prime Minister’s announcement on Wednesday of a new AI Safety Institute to explore the risks of AI developments is a very welcome step forward, as is the announced funding for a super computer and quantum computers which will be available to businesses. However, UK corporate boards must get a move on to prepare for the challenges ahead.
Developing effective governance for AI will be fundamental. The primary objective must be to ensure adoption of AI marked by consistency, openness, accountability, and transparency.
CGIUKI research earlier this year showed that only 13% of FTSE 350 companies had implemented or begun to implement policies and processes for the ethical use of AI. Two thirds of FTSE 350 boards had yet to discuss such policies, with a quarter stating they did not see a need to.
Peter Swabey, Policy & Research Director CGIUKI said, “As AI rapidly evolves, boards will need to be agile and adaptable in their approach to managing AI risks and opportunities. Boards which do not grasp this could fall behind their competitors or damage their reputation through the lack of a coherent approach.
Boards will need to develop a governance framework for AI that sets out clear roles and responsibilities, as well as policies and procedures for managing AI risks and opportunities. This framework should be regularly reviewed and updated to reflect changes in the business and the AI landscape.”
New company reporting requirements must be kept under government review to ensure there are proportional and balanced. The current government approach of charging multiple regulators to oversee AI and all developing their own processes to do so risks taking further time away from the strategic discussions and decision making needed at board level. Over 80% of FTSE 350 companies are already struggling with increasing reporting requirements.
As a minimum, boards should consider that:
- AI systems must be transparent and accountable so that it is possible for internal decision makers as well as stakeholders and regulators to understand how they have been utilised and factored into company decisions.
- Companies must recognise the potential for bias within AI systems and establish checks and reviews to audit and mitigate this risk to ensure that their systems are fair and non-discriminatory.
- Data governance policies and procedures should be kept under review to protect the privacy and security of the data collected and processed and provide assurances that it is used in a responsible and ethical manner. Boards will also need to ensure that their organisations have access to the high-quality data that is needed to train and deploy AI systems effectively.
The UK is in a reasonable place to harness the opportunities of artificial intelligence, however UK business is still some way behind the USA and China. The Prime Minister’s announcements will support them to catch-up. Corporate boards, however, must prepare now if they are to ensure arrangements are in place to take advantage of the developing technology and overcome any issues arising from its use.
