The Irish Times view on Artificial Intelligence: time for a pause

It is not very often that an industry calls for greater regulation and gets pushback from the government, but that is what is happening with artificial intelligence in the US. How this stand-off is resolved will have potentially far-reaching implications.

Dario Amodei, the CEO of Anthropic, recently published a 3,800-word letter calling for a slowdown in the development of AI until appropriate safety standards and a regulatory regime are in place. He was joined in urging new regulation by Elon Musk and Sam Altman, chief executive of OpenAI, among others.

Questions have been raised about the motivation behind Amodei’s intervention. The cynical interpretation offered by other AI companies is that introducing regulation now would stymie competition and create an oligopoly for Anthropic, OpenAI and the other behemoths in the sector. There could be some merit in this position, but the warnings about the potential dangers of the technology cannot be ignored.

These were highlighted by Jacob Coxon, formerly a senior researcher at Anthropic. He quit his role due to what he called the “out of control” nature of AI and the existential threat it poses. He claimed that there is a 10 per cent chance that AI could wipe out humanity in the next decade.

However, there has been implacable resistance from the Trump administration to pleas for regulation. In a social media post, US president Donald Trump described AI risks as a hoax and added that there was what he called a “sick conspiracy against AI and data centres”. He further added that the only guardrail needed was a “strong and smart (high IQ!) president”.

Trump’s position is shaped by considerations other than potential risks. Speaking to reporters on his way back to Washington from Ireland, he said the US was in a race with China for AI supremacy, and “whoever wins, wins”.

Moreover, the US economy – and its stock market – has become a one-way bet on AI, so any pause in its development could have significant knock-on consequences. Indeed, there was a sharp sell-off in stocks immediately following the release of Amodei’s letter.

Amodei has called for independent auditors to check AI companies for safety standards and urged global cooperation in creating rules about its development.

This would be a good starting point, but the industry cannot define its own regulatory regime. And achieving global cooperation will be difficult, with China also hostile to any restraints.

AI has many potential benefits but the concern is that its risks are unknown and unquantifiable. That is why there must be a pause in AI’s development until there is global agreement on a sufficiently robust regulatory and supervisory framework.