The Japan Times - UK retains early access to test latest US AI models

Concerns about AI safety have escalated in recent months after models developed by OpenAI and Anthropic were involved in security incidents during testing, piling pressure on governments to regulate the technology.

"We have been working closely with the US government... to retain access for Britain's AI Security Institute to the most powerful AI models before they are released," AI minister Kanishka Narayan said in a speech in London.

It marks a reversal from the previous position of President Donald Trump's administration, which had asked OpenAI and Anthropic not to share their models with the UK testing lab before they have been assessed by US authorities, according to several media outlets.

Narayan said Britain's AI Security Institute, established in 2023, had been granted access to Google's most powerful AI model, Gemini 4 Argon, which has not been released publicly over safety concerns.

Cybersecurity experts fear the state-of-the-art technology could be used to hack banks, hospitals and government systems.

Last month it emerged that a rogue OpenAI model bypassed safeguards during training and hacked an Australian government health statistics website.

Narayan on Monday warned that "it seems likely, if not inevitable, that our own cyber defences will be breached".

He urged AI labs to "listen when your scientists warn of severe dangers, investigate unexpected behaviours, and if you can't adequately control the risks, do not release the system".

To strengthen the country's defences against cyberattacks, government pilot schemes have been set up to use frontier AI to identify and address vulnerabilities in public services.

Talking to the BBC over the weekend, Microsoft president Brad Smith said the UK was best placed to lead on testing artificial intelligence models.

"In the world of technology... most people that I know would say, if you ask them, name the country that is the leader in having this technical capability.

"They would say, it's very simple. It's the United Kingdom. They are the best," Smith told the broadcaster.

He also argued for stronger safety rules, telling the broadcaster that milk and washing machines are more heavily regulated than AI.

T.Shimizu--JT