Sir Ken McCallum, the head of MI5, the UK’s domestic security service, used his annual threat update on 13 October 2026 to set out how artificial intelligence is changing national security. He said the misuse of AI he worries about most is its potential to make biological weapons more dangerous, and he announced a new MI5-hosted team to assess AI threats.
This story covers only the AI part of the speech, which MI5 published in full.
“I’m not easy to alarm”
McCallum began the AI section by stressing his caution. After thirty years of weighing threats, he said, “I’m not easy to alarm, and neither is my organisation.” But he said that “just as the benefits of AI are real, so are the national security risks.”
How AI is already being misused
McCallum listed harms that are happening now:
- “Today AI is already supercharging the cybersecurity contest between attackers and defenders.”
- AI “can produce propaganda and radicalising material tailored to the individual.”
- “We’ve caught terrorists seeking to exploit AI to improve their reconnaissance and attack methodologies, with AI-enabled drones adding a new twist.”
The risk he ranks highest
He then named a threat that has not happened yet: “Not yet realised – but the misuse case to which I’d attach greatest weight – is AI’s potential to change the stakes on existing or even new bioweapons.”
In plain terms, the concern is that AI could help people with harmful aims make existing biological weapons work better, or design new ones.
McCallum said MI5, GCHQ and MI6 are working closely with the UK’s AI Security Institute on these risks.
A new AI Threat Assessment Centre
“Today I can announce that in 2026 we’ve launched the AI Threat Assessment Centre, or AITAC,” McCallum said. “Housed in MI5 but drawing together expertise from a range of bodies including AISI, NCA and NCSC, this new team is producing all-source assessments which separate the bluster from the realities.”
AISI is the AI Security Institute, the NCA is the National Crime Agency, and the NCSC is the National Cyber Security Centre. “All-source” means the assessments draw on every kind of intelligence available, not just one stream. McCallum compared the new team to earlier units MI5 set up to analyse new dangers, such as the Joint Terrorism Analysis Centre created after 9/11.
Risks from AI itself
McCallum also spoke about what researchers call the alignment problem: the risk that advanced AI systems may not reliably act as people intend. “Put at its simplest, human history has given us no experience of facing an actor which is smarter than we are,” he said. He called this “genuinely novel” and said it deserves serious attention, “including in MI5.”
What this means
- Bioweapons top MI5’s list of AI misuse worries, even though McCallum said that risk is not yet realised.
- AI threat assessment now has a dedicated home. AITAC brings together MI5 and other UK agencies to judge which AI threats are real.
- Security chiefs are taking alignment seriously. It is unusual for the head of MI5 to discuss the risk of AI systems smarter than people.
Sources
- MI5, “MI5 Director General Delivers Annual Threat Update”, 13 October 2026. https://www.mi5.gov.uk/mi5-director-general-delivers-annual-threat-update
