Artificial intelligence (AI) chatbots are no longer being used only to spread extremist propaganda online. A new study has found that some terrorist groups, including Al Qaeda, Islamic State (ISIS), and Boko Haram, are also using AI tools to help plan attacks and make explosives.

The study, conducted by researcher Dr Juelich and shared with The New York Times before its publication, found that terrorist groups are using AI tools to design explosives, upgrade weapons and come up with ideas for attacking their enemies.

“You type in the question or use your voice and it gives you a detailed answer, like ‘How can I build a bomb?,’ and then it tells you how. It is like a human robot! We used it a lot,” one former commander in Islamic State West Africa Province, the main faction of Boko Haram, told Dr Juelich last year about using an AI chatbot.

The findings are based on nearly 60 interviews that Dr Juelich conducted over the past year with 27 former members of Boko Haram in Nigeria. Several of them said AI chatbots were being used to obtain technical information that could help improve weapons and plan attacks.

Former members said the group held training sessions, reportedly led by members linked to the Islamic State (ISIS), to teach them how to use AI chatbots more effectively.

During these training sessions, members used laptops with VPNs and encrypted software to hide their online activities. Trainers also taught them how to create AI accounts, ask questions in the right way to get useful answers and bypass the chatbots’ safety restrictions.

“The terrorists are not waiting for us to make AI safe. Their use of AI had been significantly underestimated in both scope and character,” she said.

The report pointed to the case of a 27-year-old Tunisian man who was arrested in May over an alleged plot to attack a museum or a Jewish site in Paris. Investigators said AI was used to help plan the attack.

Daniel Byman, a terrorism expert at Georgetown University, said terrorist groups do not rely on just one chatbot. Instead, they switch between different AI platforms, including ChatGPT, Claude, Gemini, Grok and DeepSeek.

CIA Director John Ratcliffe recently compared terrorist groups’ plans to use AI to “digital nuclear weapons”.

US intelligence officials say some terrorist groups are also using AI to help produce 3D-printed weapon parts. According to a former US official, AI is also being used to design and manufacture drone parts, weapon components and ammunition fittings.