The report titled ‘God has helped us, and so will AI: How the Terrorist Group Boko Haram Uses Frontier AI’ is based primarily on 57 interviews with 27 former Boko Haram members.
A new study published by the University of Cambridge has stated that some Boko Haram fighters used frontier artificial intelligence (AI) tools, including ChatGPT, Claude, Gemini, Grok, Meta AI and DeepSeek, to plan attacks, design explosive devices and improve battlefield operations.
The study, authored by Antonia Juelich, a terrorism researcher at the Cambridge Programme on AI Science & Policy (CASP), presented what it calls the first field-based evidence of AI adoption by terrorist organisations.
The study was first reported by The New York Times. A review of the 93-page report by PREMIUM TIMES found that many of its central claims could not be independently verified and are based largely on interviews with former Boko Haram members who defected from the group.
The researcher acknowledged these limitations in the study, noting that several claims could not be independently corroborated because of the secrecy surrounding Boko Haram and the difficulty of accessing active members.
The report also provides no forensic evidence, platform records or technical data linking Boko Haram directly to the AI systems identified, nor does it indicate whether the companies behind those products were contacted for comment or independently confirmed attempts by the group to use their services.
The researcher further cautioned that the study cannot determine whether AI measurably improved Boko Haram’s operational capabilities, saying it instead documents former members’ perceptions that the technology made them more effective.
Now in its 17th year, the Boko Haram insurgency has killed more than 35,000 people, displaced over two million others and devastated Nigeria’s North-east, while increasingly spreading to other parts of the country despite years of military campaigns to contain it.
Unlike previous studies that focused mainly on extremist propaganda, the report states that Boko Haram’s two factions—the Islamic State West Africa Province (ISWAP) and Jama’atu Ahlis Sunna Lidda’awati wal-Jihad (JAS)—used frontier AI across combat operations.
According to former members interviewed for the study, AI tools were used to plan attacks, improve operational security, troubleshoot weapons, provide logistical advice and assist in designing improvised explosive devices. Some respondents also claimed AI helped commanders develop battlefield tactics and improve drone weaponisation.
“We mostly used it in three ways: the first one is to learn how to assemble and use guns and how to manufacture bombs,” the researcher quoted a former ISWAP commander as saying. “The second one is for surveillance, like how to improve our surveillance strategies to monitor what is happening in our camps and also to better understand our enemy and prepare attacks. The third one is to make plans, like, when we come up with new ideas on how to attack, we ask it for tactics on how to make it work in practice to be successful.”
The researcher further stated that the insurgents learned methods for bypassing AI safety guardrails through prompting techniques taught by foreign trainers, allowing them to obtain responses that commercial AI systems are designed to restrict.
One of the study’s most consequential claims is that foreign Islamic State fighters allegedly introduced AI to ISWAP beginning around 2023.
“The white guys came and taught us,” Ms Juelich quoted a former ISWAP mid-ranking commander, further clarifying that he was referring to operatives from Libya, France, and Arab countries with a lighter complexion than their own.
Former commanders interviewed by the researcher said foreign trainers supplied laptops, virtual private networks (VPNs), encrypted software and paid subscriptions to multiple AI platforms while conducting training sessions for selected commanders.
According to the report, dedicated AI units were established within both ISWAP and JAS to manage access to AI tools, train other fighters and support operational planning.
While the report stops short of prescribing detailed policy measures, the researcher urges governments, artificial intelligence companies and the academic community to work more closely to better understand and respond to the emerging use of AI by terrorist groups.
“The specific policy implications are beyond the scope of this paper,” the author wrote, noting that the study was intended primarily to present empirical findings that could inform future policy discussions.
According to the researcher, greater attention should be paid to how terrorist groups could exploit AI for less dramatic but operationally significant purposes, including logistics, communications, tactical planning and troubleshooting weapons systems.
The author further recommended expanded empirical research to determine whether AI adoption extends beyond Boko Haram to other Islamic State affiliates, al-Qaeda-linked groups and non-jihadist armed organisations. Such studies, the report says, are necessary to establish whether the patterns identified in this research are isolated or indicative of a broader trend in the evolution of militant organisations.
She also urged AI developers and policymakers to involve conflict and terrorism researchers more directly in AI safety assessments, arguing that technical testing alone cannot adequately capture how militant groups make operational decisions or exploit emerging technologies. She said understanding the behaviour, motivations and organisational structures of extremist groups is essential to accurately assessing AI-related security risks.
More broadly, the researcher argued that the findings should prompt governments, intelligence agencies and AI companies to reassess their assumptions about the pace at which terrorist organisations may be adopting frontier AI technologies, while stressing that determining the appropriate policy response lies beyond the scope of the study.
