The UK AI Safety Summit – How can the UK show global leadership in regulating the AI sector?

By Toby Varian, Account Manager-Tech & Finance at Sovereign Strategy

There can be no doubt that AI is one of the most existential issues facing society and the defining technological advancement of the century. According to the UK Government, the UK AI sector is estimated to be worth a trillion dollars by 2035, which is what our entire tech sector is worth today. However, the Government has also identified a range of tangible risks that AI poses to society, including AI-generated disinformation to undermine elections and mass disruption to the job market. Across the world, politicians are now recognising the importance for an internationally coordinated response to harness the impact these technologies will have on us all.??

The Prime Minister, Rishi Sunak has long been viewed as a tech optimist and has outlined plans to make the UK the intellectual and geographical home of global safety and AI regulation. To this effect, the UK Government will be hosting the very first Global AI Safety Summit this week (1 and 2 November 2023) at Bletchley Park, hosting world leaders, tech experts, industry leaders and civil societies. The summit will focus on navigating the risks of “frontier AI” which include task-related and large language models. Commenting on the Summit in October, Sunak remarked that the international community is at a crossroads with regulating AI, and failure to take decisive action could lead to AI becoming too advanced without having the proper safety systems and regulations to control it.??

The significance of the summit and the need for greater clarity and cooperation from world leaders should not be underestimated. There are several key questions that should be addressed by governments and policy makers at the summit and beyond.??

1 - What would an international framework look like and how would it affect businesses???

We have already begun to see countries take different approaches to regulating AI. The EU is moments away from delivering their AI Act where AI models and tech companies in scope of the legislation must comply with a strict set of rules and governance from Brussels. Similarly, the US launched a new executive order on AI on 30 October 2023, requiring new safety assessments and security standards on AI and an assessment of how it will impact the labour market. Elsewhere, China has taken a stricter approach to AI legislation, focusing primarily on digital platforms and deepfake technologies.???

The UK Government believes that a unified approach to AI regulation would help to better manage cross border risks and regulatory gaps across different markets. This approach must work not just for international partners, but also for the companies and regulators who operate within these regions who will have more regulatory burdens and uncertainties to deal with. Many of these companies will be concerned that an overly cautious approach to regulating AI will stifle innovation and technological developments that they could use to tackle the same issues that politicians are currently concerned about.??

2 - How will the summit address concerns for human safety???

One of the main concerns that many people have with AI is the risk that these new technologies could cause mass layoffs across different sectors. There have also been instances where bias in AI systems has led to discrimination in job recruitment, financial lending and healthcare. Being too lenient towards the rollout of AI could widen existing societal inequalities and make more people sceptical about the benefits of a digital revolution. Any outcomes from the summit should address how AI can be used to improve productivity within the workforce and society at large and not at the expense or safety of humans.???

3 - What will the legacy of the summit be???

The summit is a critical opportunity for Sunak to establish a legacy project as he faces significant challenges in next year’s General Election, which begs the question as to whether the UK is really in the best position to lead the conversation on AI regulation. If the Labour Party (who are currently 15 points ahead in the election polls) form the next incoming Government, they will have to act fast to set out their long-term AI policies and decide if they want to continue supporting the findings made in the summit.???

In a year where so many world leaders have made proposals to regulate AI, and another global summit is already scheduled in New Delhi next month with more world leaders and experts expected to attend, it is very possible that the outcomes and findings of the UK Summit could be lost in an unrelenting media cycle that will soon focus on issues that are more important to the British public ahead of an election. Therefore, the Summit ought to give UK regulators and tech companies some indication as to what approach the Government will take towards its own policy making in the final parliamentary session before an election.???

Ultimately, the findings of the summit will be published after the summit, which along with the upcoming Kings Speech on 7 November 2023 suggests that the next fortnight will be significant for understanding the direction of travel for AI both domestically and abroad.??