We could all die. I don’t just mean at the end of our lives. I mean as a species. That’s the threat posed by artificial intelligence, according to more than 350 scientists and industry leaders who have written a joint statement. It reads, in full: “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”
It’s at times like these that people turn to their government for some reassurance. Surely, they’ve got it covered?
The British government is keenly light touch. In March it said it would pursue a “pro-innovation” approach to regulation. In practice, this means it’s cautious about putting industry rules on a statutory footing – ie, giving it the force of law – or setting up a new regulator because of fears they would be stifling. Instead, the government will ask pre-existing regulators such as the Health and Safety Executive and the Equality and Human Rights Commission to incorporate AI regulation into their remits. One concern is this approach won’t give regulators the required power to curtail the dangers of AI.
Another is that the UK isn’t keeping up with the US and the EU. The Europeans are in the final stages of an AI Act, which is being celebrated as the most comprehensive set of regulations to date. The government would say tighter regulation in the EU is an opportunity for the UK to attract AI businesses. But the risk is that the size of the EU market means it could be cheaper for companies to comply with their rules everywhere rather than accept different, fewer rules in the UK – neutering Britain’s competitive advantage. Even that depends on how stringent the final version of the EU’s AI Act is.
As for the US, British overtures to the White House to create something similar to the US-EU Trade and Technology Council have been unsuccessful. Much as with the green technology subsidies in America’s Inflation Reduction Act last year, the danger for the UK is that it gets squeezed between the two behemoths.
Regardless, uncertainty lies at the core of the AI debate. That the joint statement was a single sentence – perhaps the only way to get everyone to sign – reflects the divergence of opinion on where the technology, and therefore also the regulation, is heading.
This piece first appeared in the Morning Call newsletter; subscribe to it on Substack here.