AI as a New Weapon for Terrorists? Recent Findings Reveal Growing Threat
A new report by Tech Against Terrorism, a cyber monitoring organisation, has highlighted the growing utility of artificial intelligence (AI) for would-be extremists. Researchers submitted over 2,300 prompts, drawn from real terrorist use cases, to 27 different Large Language Models (LLMs). The findings revealed that 32% of the queries yielded information that was ‘genuinely useful’ for malicious purposes. When the same questions were framed as research requests, the success rate climbed to 42%.
Whilst extremist groups like the so-called Islamic State (ISIS) and al-Qaeda have used AI for propaganda and meme generation for several years, the technology’s role is rapidly expanding. Analysts from Militant Wire noted a significant increase throughout 2025 in incidents where terrorists used AI tools to plan, research, and prepare for attacks. This shift marks a move beyond content creation towards operational support.
Rueben Dass, a researcher at the S. Rajaratnam School of International Studies in Singapore, told DW that AI chatbots are beginning to fulfil the role of ‘virtual planners’. Previously, lone actors were often guided by remote human handlers online. ‘I wouldn’t say that humans have been completely replaced, but to some extent, these lone actors are now turning to AI, like ChatGPT, for that support,’ Dass explained. The trend is also evident among groups like Boko Haram, whose members have reportedly used models such as ChatGPT, Claude, and Gemini to plan attacks and maintain operational security.
Moustafa Ayad, Executive Director at the Institute for Strategic Dialogue (ISD), noted that the jihadist ecosystem is leveraging AI across a wide spectrum. This ranges from creating TikTok dance memes to sophisticated cross-border propaganda and dedicated ‘jailbreaking’ efforts to bypass safety guardrails for operational planning. ‘That’s the problem,’ Ayad stated. ‘AI can simultaneously accelerate propaganda and support attack preparation.’