By Ghazali Ibrahim
A new report by The New York Times has revealed that Boko Haram and Islamic State West Africa Province (ISWAP) fighters used artificial intelligence (AI) chatbots to improve bomb-making techniques and refine aspects of their military operations.
The report, based on interviews conducted by Antonia Juelich, a terrorism and technology researcher at the University of Cambridge, said former insurgents disclosed that they relied on AI platforms including ChatGPT, Claude, Gemini, Grok and DeepSeek to obtain technical information.
According to the report, one former Boko Haram commander said the group turned to AI after an attack on a military base failed because a defensive trench prevented their motorcycles from advancing.
The commander said the fighters asked AI tools how to modify their motorcycles to overcome such obstacles.
The chatbots reportedly provided guidance on adjustments to improve speed and acceleration, enabling the group to train for similar missions.
“We saw in a movie how motorcycles can jump over bridges. We used AI to learn how to do this. We gave it information, like what motorcycles we use and the distance we need to jump, and it gave us steps on what we have to do,” the former commander was quoted as saying.
The report also cited a former ISWAP commander who claimed the group used AI to seek information on explosive devices.
According to him, the technology reduced the risks associated with trial-and-error experimentation by providing detailed responses to technical questions.
The report said some former insurgents claimed AI-assisted information helped them increase the destructive power of improvised explosive devices by identifying combinations of chemicals that produced stronger explosions.
Responding to the findings, OpenAI said the reported use of its services by terrorists violated its usage policies.
Google and Anthropic also stated that their AI systems include safeguards designed to reject requests for dangerous or harmful information.
However, the former fighters told the researcher they were sometimes able to bypass those restrictions by disguising their requests as legitimate academic or technical inquiries.
