The head of Britain’s MI5 spy agency has raised concerns about the future risks of AI systems operating without human control. Ken McCallum spoke on Thursday about how artificial intelligence is already being used by security services, terrorists, and government actors, but warned that autonomous AI could one day pose a serious security threat.
In his yearly address on threats to the UK, McCallum explained that MI5 uses AI to improve its work. However, he noted terrorists also use AI for spreading propaganda and scouting targets, while some governments exploit it to interfere with elections and carry out cyberattacks.
Looking ahead to 2025 and beyond, McCallum said it’s important to prepare for “non-human, autonomous AI systems” that might escape human supervision. He was careful to avoid dramatic warnings and said he’s not predicting sci-fi style dangers. Still, he stressed that as AI keeps improving, security agencies need to think seriously about how to protect the country.
“Artificial intelligence may never ‘mean’ us harm,” McCallum said, “but it would be reckless to ignore the potential for it to cause harm.” He assured the public that MI5 is actively monitoring and addressing these challenges.
This comes amid growing concerns from tech leaders about the risks of unchecked AI, including fears of scenarios like those depicted in the Terminator movies, where self-aware machines threaten humanity. McCallum’s message strikes a balance between caution and reassurance, highlighting the fine line between hype and real danger in the AI landscape.
As AI becomes more powerful, the conversation around its impact on security will only grow. For now, Britain’s intelligence services are keeping a close eye, ready to respond if these emerging risks start to turn into real threats.