MI5 chief calls Russia 'aggressor number one', warns Iran is targeting UK bases

The head of Britain's domestic intelligence service, MI5 Director General Ken McCallum, said on Tuesday, 13 October 2026 that Russia remains the United Kingdom's "aggressor number one", that Iran has made a major shift towards targeting British military sites, and that artificial intelligence is making hostile actors more dangerous. In his annual threat update, delivered at MI5's London headquarters, Thames House, he said the agency now devotes more resources to countering hostile states than it did during the Cold War. He also announced a new AI Threat Assessment Centre inside MI5.

The annual update is one of the few times the head of a Western intelligence service sets out publicly how the threat picture is changing. This year's message is that the line between peace and conflict has blurred. McCallum said, "The Russian military and intelligence services are once again escalating their attempts at hybrid warfare, being careful to stay below the NATO Article 5 threshold, which would be foolish for them to cross." Article 5 is the clause that treats an attack on one NATO member as an attack on all. Hybrid warfare, which includes sabotage, cyber attacks, disinformation and the use of hired criminals, is designed to cause damage without triggering that response.

The warning on Iran is significant because it suggests Tehran is moving from targeting dissidents and journalists towards harder targets such as military bases. More than 50 arrests in a year shows the scale of the activity British agencies are dealing with. For Europe, the speech adds to a pattern: Germany and Estonia have also accused Moscow of hybrid attacks in recent months.

The AI section of the speech may have the widest relevance. Security services everywhere are trying to work out how cheap, powerful AI tools change what small groups can do. McCallum said MI5 had caught terrorists trying to use AI to plan attacks, and that his greatest worry is extremists using AI to help develop biological weapons, a threat he said has not yet been realised. A dedicated centre to assess such risks is an early model other governments, including India's, are likely to study as they build their own AI safety and security structures.

Short-term: The speech is likely to feed pressure in Britain for more counter-intelligence funding and closer monitoring of Iran- and Russia-linked networks, including criminals hired as proxies.

Long-term: The new AI centre could shape how Britain and its allies assess AI-enabled threats, from drone attacks to biological risks, and how they share that work across agencies.

Who is affected: British security agencies and armed forces, NATO allies facing Russian hybrid activity, Iranian dissidents abroad, technology companies whose AI tools may be misused, and governments designing AI security policy.

Questions and Answers