UK Spy Chiefs Call AI One of Britain's Biggest Security Threats

The people whose job is to worry about Britain's security have put AI near the top of the list. According to The Times, which reported the story on October 11, the heads of MI5, MI6 and GCHQ told ministers that advanced AI now ranks among the country's biggest national security threats.

The warning came at a Cabinet briefing on September 8 from Ken McCallum of MI5, Blaise Metreweli of MI6 and Anne Keast-Butler of GCHQ. The news is the report of what was said, not a fresh alarm; the meeting was described as part of a routine assessment of the threat picture rather than a response to any imminent danger.

What they told ministers

The concern, as relayed by the outlets summarising the briefing, is that hostile states, terrorist groups and lone actors could use increasingly capable AI to mount cyberattacks and cause serious harm. Russia and Iran were named as the states most likely to weaponise the technology. The risk was not limited to hacking; the chiefs reportedly pointed to propaganda and interference as well. MI5's director general is said to have told the room that AI was "the biggest thing you should be thinking about."

At least one account reports a minister comparing the mood of the meeting to the early days of the Covid pandemic and the 2008 financial crisis, moments when the warnings were loud and the response, for a while, was not. Whether that comparison holds is a matter of record yet to be written.

A caveat worth keeping

I would hold these details a little loosely. The original Times report sits behind a paywall, and the specifics here come from secondary outlets summarising it, so the exact wording may differ from the source. The broad shape is consistent across the coverage: three intelligence agencies, one Cabinet briefing, and a message that the technology is now a first-order security concern.

It lands in familiar company. Anthropic has already warned its own investors of catastrophic risk, and the labs themselves are the ones building the capability the spy chiefs are worried about. When the people who sell the technology and the people paid to defend against it start describing the danger in similar terms, the warning is worth hearing twice.