Government officials from the UK have been touring the globe trying to convince world leaders to subscribe to its light-touch approach to AI regulation. It is also hoping to convince countries to agree to specific statements on AI risks at next month’s Bletchley Summit. The UK hopes to make itself a global leader in AI safety, but its efforts may be coming too late as Japan and the EU have started to align on a different approach to policing artificial intelligence.
While the UK is a key player in the global technology race, with an AI market worth more than £16.9bn and investment in the billions over the next decade, every other nation is also investing heavily in the game-changing technology. Globally, the largest nations are also debating ways to handle legislating and regulating AI.
Leaders of the G7 nations have been meeting to discuss and establish international AI regulations and these could be in place before the end of the year, Japanese Prime Minister Fumia Kishida said earlier today. The wealthiest nations are working on a coordinated approach known as the Hiroshima AI process which will include a code of conduct for nations building advanced AI systems. Made up of the US, Canada, Japan, the UK, France, Germany and Italy, it will also cover governance, IP rights, information and responsible use.
Independently, the EU has been lobbying other nations, including Japan, to adopt its more prescriptive approach to the regulation of AI. Unlike the UK which has a light-touch approach, the EU’s comprehensive AI act includes restrictions on the use of the technology, requirements around data used in training generative AI and outright bans on some forms of biometric AI such as police use of facial recognition.
Now the EU says Japan had begun to more closely align with its approach. Vera Jourova, the European Commission’s vice-president for values and transparency, told Reuters: “I see a lot of convergence in how we look at AI and generative AI.” This could leave the UK scrabbling for allies, forced to adapt to a more prescriptive approach, or out in the cold.
The government has pledged to fund UK AI development, with a £900m investment set aside for AI and exascale computing, and money for a foundation model task force to research AI safety and ways to utilise the technology for public services.
Its hoped the summit will be an opportunity for the UK to convince other leading nations, and companies, to agree to its global approach. Co-organiser Matt Clifford wrote on X: “Our view is that most countries will want to develop their own capabilities in this space, particularly to evaluate frontier models.” He added that the UK frontier AI taskforce is building expertise and capability in AI safety and that that information could be shared with others.
“Collaboration is key to ensuring we can manage risks from frontier AI – with civil society, academics, technical experts and other countries,” Clifford wrote. “That’s the positive discussion we are looking forward to at Bletchley Park in November and we’re optimistic about the potential impact.”
But the summit has been criticised for its narrow focus on next-generation frontier models and limited delegate list, which includes only those company CEOs building frontier AI models. This includes the likes of Microsoft, Google, ChatGPT-maker OpenAI and Anthropic.
UK in the last chance saloon on AI safety?
The upcoming summit could be the UK’s last opportunity to play a leading role in shaping global regulation of AI according to Alistair Dent, chief strategy officer at data company Profusion. He says that other nations are piecing together a “risk-based” approach to regulation that is focused on minimising the risk for human harm, but the UK appears committed to a light-touch, pro-innovation approach that leaves it out of step globally. This, says Dent, will lead to an increased risk of AI being used in undesirable ways and causing the public to lose confidence, which will in turn cause the government to be forced to over-correct.
“AI is developing much faster than any regulation can keep up with so seeking to address individual applications or products is likely to be ineffective,” Dent says. He notes that the UK policy debate makes little mention of the need for an ethical framework around the use of AI. Such a framework could form the basis of flexible regulation that prevents undesirable outcomes for the end user. “Ideally, these principles would be enforced globally and the AI summit is an excellent place to start this discussion,” he says. “However, the reality is that just 100 attendees is hardly the quota needed to ensure the holistic, rounded discussion.”
And while the UK’s AI investments are welcome, Claire Trachet, CEO of business advisory company Trachet, says it is “essential that the UK continues to balance its drive for innovation with creating effective regulation that will not stifle the country’s growth prospects.” She explains: “While the UK possesses the potential to be a frontrunner in the global tech race, concerted efforts are needed to strengthen the country’s position.
“By investing in research, securing supply chains, promoting collaboration, and nurturing local talent, the UK can position itself as a prominent player in shaping the future of AI-driven technologies.”
Read more: Concern as AI Safety Summit ‘limited to 100 delegates’
