In his annual address to MI5 staff, Director General Sir Ken McCallum painted a stark picture of the security landscape facing the United Kingdom. He argued that the combined pressure from Russia, Iran and emerging artificial‑intelligence (AI) technologies now outweighs the threats that dominated the Cold War era.
State threats on the rise
Sir Ken said that threats from foreign adversaries have quadrupled in the last decade, with Russia identified as the "aggressor number one". He described Russia’s strategy as a campaign of hybrid warfare – a mix of cyber‑attacks, sabotage and disinformation – aimed at unsettling European nations while avoiding a conventional clash with NATO. The director highlighted that, despite the heavy cost of its war in Ukraine, Russia continues to pursue covert operations, noting that the Kremlin’s “global toolkit is weak” after nearly 1,000 expulsions of Russian spies from the UK.
Iran was also singled out for expanding a hybrid campaign on British soil. Sir Ken cited 50 arrests over the past year linked to Iranian‑sponsored activity, targeting Jewish communities and military sites. He referenced a recent incident at RAF Fairford in Gloucestershire, where three suspicious vans prompted a police operation resulting in seven arrests and bail decisions.
Extremist terrorism and the AI factor
The director warned that the threat from extremist groups is now "pretty much of equal scale" to state‑backed threats. He highlighted a rise in extreme right‑wing terrorism, noting that many cases involve young individuals radicalised online. Sir Ken recounted a recent investigation of a 13‑year‑old extremist, which generated intelligence that helped a foreign partner prevent a lethal attack.
AI, according to Sir Ken, is "supercharging the cyber‑security contest between attackers and defenders". He described several ways AI is being weaponised:
- Generating propaganda and radicalising material tailored to individuals.
- Enabling terrorists to improve reconnaissance and attack methods, including the use of AI‑enhanced drones.
- Potentially facilitating the development of bioweapons, a risk he said is "not yet realised" but could "change the stakes".
To address these emerging dangers, MI5 has launched an AI Threat Assessment Centre tasked with separating hype from genuine risk. Sir Ken emphasized collaboration with the UK’s AI Security Institute and the broader intelligence community (MI5, GCHQ, MI6) to mitigate AI‑related threats.
Legal tools and broader context
The director noted that the National Security Act – a 2023 law designed to target individuals working against UK interests for foreign powers – is now being used heavily against threats from Russia, Iran, China and extremist groups.
He also referenced the recent elevation of the UK’s terrorism threat level to "severe" for the first time in five years, underscoring the heightened likelihood of an attack.
Analysis
Sir Ken’s briefing underscores a strategic shift within MI5: resources are being reallocated from traditional espionage to a broader spectrum of hybrid and AI‑driven threats. The emphasis on AI reflects a growing consensus among Western security agencies that generative technologies could accelerate both cyber‑offensive capabilities and the creation of tailored extremist content. By establishing a dedicated AI assessment unit, MI5 signals its intent to stay ahead of a rapidly evolving threat environment.
Russia’s continued reliance on covert influence operations, despite a weakened spy network, suggests a pivot toward more deniable tactics such as cyber‑intrusions and propaganda campaigns. Iran’s focus on community‑level intimidation, particularly against Jewish institutions, aligns with its broader pattern of leveraging diaspora networks to exert pressure abroad.
Uncertainty
While the director highlighted the potential for AI‑enabled bioweapons, concrete evidence of such capabilities being operationalised by extremist groups remains limited. The exact scale of AI‑driven radicalisation – especially the effectiveness of AI‑generated propaganda – is still being measured, and the long‑term impact on terrorist recruitment is uncertain.
Furthermore, the precise extent of Russia’s hybrid activities in the UK, beyond publicly disclosed expulsions and alleged disinformation campaigns, cannot be fully quantified from open‑source information.
Overall, the address paints a picture of a security apparatus adapting to a complex mix of traditional state aggression, home‑grown extremism, and the nascent challenges posed by artificial intelligence.