The Pulse
MI5 Launches AI Threat Assessment Centre With Three Partner Bodies
MI5 has launched the AI Threat Assessment Centre, a unit based at the agency that draws on expertise from the AI Security Institute, National Crime Agency and National Cyber Security Centre. Director General Ken McCallum says it is producin
AI.info Team · Written with AI ·
MI5’s new unit draws on three partner bodies
Three outside organizations contribute expertise to MI5’s new AI Threat Assessment Centre: the AI Security Institute, the National Crime Agency and the National Cyber Security Centre. MI5 Director General Sir Ken McCallum announced the centre’s launch on October 13, 2026, during his annual threat update at Thames House in London. Based inside MI5, the centre is producing all-source assessments of national security threats involving artificial intelligence.
McCallum said the assessments are meant to distinguish actual risks from inflated claims as government weighs AI’s benefits against possible misuse. The centre draws together expertise across organizations rather than operating as a standalone research lab, according to MI5’s announcement. The agency says AITAC will provide assessments to a range of government departments.
“Today I can announce that in 2026 we’ve launched the AI Threat Assessment Centre, or AITAC.”
Sir Ken McCallum, Director General, MI5
MI5 names terrorist misuse and bioweapons risks
McCallum linked the centre’s work to several distinct security concerns. He said AI is already intensifying the cybersecurity contest between attackers and defenders, and can generate propaganda tailored to individuals. He also said MI5 has caught terrorists seeking to use AI to improve reconnaissance and attack methods, including alongside AI-enabled drones.
McCallum singled out bioweapons as the misuse scenario he weighs most heavily, while making clear it has not yet materialized. He said AI could change the stakes around existing or new biological weapons. The UK intelligence community—MI5, GCHQ and MI6—is working with the AI Security Institute on these risks, he said.
The centre also considers risks from AI itself
AITAC’s remit sits alongside a separate concern in McCallum’s speech: possible risks from AI systems themselves, including what researchers call the alignment problem. That is distinct from people using AI to aid cyberattacks, influence campaigns or physical violence. McCallum described the prospect of an actor capable of outthinking humans as novel and deserving serious attention.
MI5 presented the new unit as part of a longer pattern of creating joint assessment bodies for emerging security threats. McCallum compared AITAC with the Joint Terrorism Analysis Centre, formed in the period after the September 11 attacks, and the Joint State Threats Assessment Team, established about a decade ago. He said those bodies were built to improve analysis of dangers that crossed organizational boundaries.
MI5 has not disclosed AITAC’s size or budget
The announcement describes AITAC’s location, partners and assessment role, but does not state how many people work there, what budget it has or how often it will issue assessments. Nor does it explain how the centre will share its findings with departments or what form those assessments will take. Those operational details remain absent from MI5’s public description of the new centre.