MI5 looking at potential risk from out-of-control AI

British spies are working to combat the potential risk posed by rogue artificial intelligence (AI) systems, the head of MI5 has said.

Sir Ken McCallum said it would be “reckless” to ignore AI’s potential for harm.

Speaking at the Security Service’s Thames House headquarters, he said he was “not foreseeing Hollywood movie scenarios” but insisted intelligence agencies must consider the risks.

He said: “MI5 has spent more than a century innovating our human – and sometimes inhuman – enemies.

“But in 2025, as we deal with today’s threats, we also need to consider the next frontier: potential future risks from non-human, autonomous AI systems that can evade human surveillance and control.

“Given the risks of exaggeration and fear-mongering, I will choose my words carefully: I am not predicting Hollywood movie scripts.

“Overall, I am a technology optimist who thinks AI brings real benefits.

“But as AI capabilities continue to strengthen, you can expect organizations such as MI5 and GCHQ, as well as the UK’s groundbreaking AI Security Institute, to be thinking hard today about what defending the region might need to look like in the coming years.

“AI may never be ‘meant’ to harm us. But it would be reckless to ignore its potential to do harm.”