Sir Ken McCallum says terrorists have tried to use artificial intelligence to improve reconnaissance and attack methods. Britain is establishing a dedicated center to assess emerging AI-related security risks.
LONDON | October 15, 2026 – Britain’s domestic intelligence chief has warned that artificial intelligence could significantly increase the capabilities of terrorist groups, potentially making it easier for hostile actors to plan attacks, operate AI-enabled drones and explore the development of biological weapons. The warning comes as the United Kingdom faces a widening range of security threats linked to terrorism, hostile states and rapidly advancing technology.
Sir Ken McCallum, director-general of MI5, delivered the warning during his annual national security update at the agency’s headquarters in London on October 13. He described a changing security environment in which technological advances could give dangerous individuals access to capabilities that were previously difficult to obtain.
McCallum said MI5 had already encountered terrorists seeking to exploit AI to improve their reconnaissance and attack methods. He also highlighted the potential for AI-generated propaganda and personalized radicalization material to make extremist recruitment and influence campaigns more effective.
The intelligence chief’s remarks underline a growing concern among security agencies: AI can help legitimate researchers, businesses and governments, but the same technology may also be misused by terrorists and state-linked actors.
AI could make terrorist operations more effective
One of MI5’s concerns is that AI could help individuals and extremist groups gather information, identify potential targets and refine their operational planning. McCallum said investigators had caught terrorists attempting to use the technology to improve their reconnaissance and attack methodologies.
AI-enabled drones add another dimension to that threat. The expanding use of remotely operated and increasingly autonomous drone systems in the war in Ukraine has demonstrated how relatively accessible technology can change modern conflict.
Security officials are concerned that terrorist groups could adapt similar technologies for attacks inside Britain. AI may potentially help hostile actors interpret information, automate some tasks or improve the coordination of existing systems.
However, the warning does not mean that a major AI-directed drone attack in Britain is imminent. Rather, it reflects the intelligence service’s assessment that the technology could create new opportunities for attackers and complicate efforts to detect and prevent plots.
The broader challenge for security agencies is to understand how quickly these capabilities are evolving and how existing counterterrorism measures may need to adapt.
Biological weapons emerge as a major concern
The potential misuse of AI in biological research was among the most serious issues raised by McCallum. He warned that, if safeguards fail, the technology could make dangerous biological knowledge more accessible to individuals who lack advanced expertise.
AI systems can process large quantities of scientific information and help users understand complex research. Security officials fear that malicious actors could attempt to exploit these capabilities to explore dangerous biological applications or overcome barriers that have historically made biological weapons difficult to develop.
McCallum identified the possibility of AI changing the risks associated with existing or new biological weapons as a particularly significant concern. He also emphasized that this remains a prospective threat rather than an established pattern of successful terrorist attacks involving AI-generated pathogens.
Developing and deploying a biological weapon requires far more than obtaining information from an AI system. Practical experimentation, specialized facilities, biological materials and the ability to overcome technical and safety challenges remain substantial obstacles.
Nevertheless, intelligence agencies are concerned that advances in AI could lower some knowledge barriers and increase the capabilities of individuals or groups with malicious intentions.
The issue presents a difficult challenge for governments and technology companies. They must protect legitimate scientific research while reducing the risk that powerful tools could be used to facilitate serious harm.
Personalized propaganda and online radicalization
McCallum also warned that AI can produce propaganda and radicalizing material tailored to individual preferences. This could allow extremist groups to create more targeted messages and adapt their communications for different audiences.
Generative AI can produce text and other content quickly, potentially reducing the time and effort needed to develop large volumes of persuasive material. Malicious actors may also attempt to use the technology to imitate trusted sources, spread misleading narratives or exploit existing social divisions.
For intelligence agencies, the concern extends beyond the production of propaganda itself. Personalized content may make it harder to identify coordinated influence campaigns and assess how online activity could translate into real-world violence.
AI is also being used by security services to analyze information and support investigations. The technology therefore represents both a potential threat and a tool that agencies can use to strengthen their defensive capabilities.
Britain establishes an AI Threat Assessment Centre
In response to emerging risks, MI5 is establishing an AI Threat Assessment Centre, known as AITAC, to examine how advances in artificial intelligence could affect national security.
The initiative is intended to help authorities assess developing threats, understand how AI systems could be misused and identify measures to reduce the risks. The center is expected to work with other security and law-enforcement bodies, including the National Crime Agency and the National Cyber Security Centre, which is part of GCHQ.
The initiative reflects the view that AI-related security risks cannot be addressed by intelligence agencies alone. Government departments, technology developers, researchers and law-enforcement organizations will need to cooperate as the technology evolves.
A central task will be distinguishing realistic and emerging threats from speculative scenarios, while ensuring that authorities can respond quickly when credible risks arise.
Russia and Iran remain major security concerns
McCallum’s annual assessment also highlighted the wider geopolitical environment confronting Britain. He described Russia as the leading aggressor in the current security landscape and warned that hostile states were continuing to use a combination of cyber activity, intimidation and other forms of pressure against Britain and its allies.
Iran was another major concern. McCallum said Iranian-linked threats were increasingly drawing attention to potential targets beyond traditional dissident communities, including military installations and other sensitive sites.
The security concerns include activity involving RAF Fairford in Gloucestershire, a military base used by US forces. British authorities have investigated suspected Iran-linked activity, while Iran has denied allegations of involvement in hostile operations.
According to McCallum’s assessment, more than 50 arrests connected to Iran-related activity had been made in Britain over the preceding year. Arrests do not, by themselves, establish guilt, and individual allegations remain subject to investigation and legal proceedings.
China also remains a focus of British counterintelligence efforts, particularly over concerns about espionage and attempts to acquire sensitive information or influence strategic sectors.
These state-linked risks are developing alongside the threat posed by terrorist organizations and individuals who may be radicalized online.
Technology is reshaping national security
The MI5 warning highlights how artificial intelligence is changing the security calculations of governments around the world. Tools that can accelerate research, automate tasks and generate convincing content may also create opportunities for people seeking to commit violence or disrupt critical infrastructure.
The challenge is not simply to restrict AI development. Authorities must also understand how malicious users might exploit the technology, strengthen safeguards and ensure that intelligence services have the expertise needed to identify emerging threats.
For Britain, the creation of AITAC signals a more structured effort to assess these risks before they develop into operational threats. The effectiveness of that approach will depend on timely intelligence, cooperation with technology companies and safeguards that limit harmful uses without unnecessarily obstructing legitimate innovation.
McCallum’s message is a warning about the direction of travel rather than a prediction that every feared scenario will occur. AI has the potential to improve security and scientific progress, but its misuse could make some forms of terrorism more accessible and harder to counter.
As the technology advances, British intelligence agencies face the task of keeping pace with new capabilities while separating credible dangers from possibilities that remain uncertain.
