How Boko Haram Fighters Used AI Chatbots To Build B0mbs, Plan Att∆cks – Cambridge University Research

 

 

How Boko Haram Fighters Used AI Chatbots To Build Bombs, Plan Attacks – Cambridge University Research

According to the study, members of ISWAP and JAS were trained to use AI systems and bypass safeguards designed to prevent chatbots from responding to dangerous requests.

Boko Haram fighters have used artificial intelligence chatbots from major technology companies to obtain technical and military advice, including assistance related to explosives, combat operations and weapons, according to a research by the University of Cambridge.

Researchers interviewed 27 former members of Boko Haram’s two factions — the Islamic State West Africa Province (ISWAP) and Jama’at Ahl as-Sunnah lid-Da’wah wa’l-Jihad (JAS).

Several former fighters said they routinely used ChatGPT, Claude, Gemini, Grok, Meta AI and DeepSeek before, during and after attacks.

One former ISWAP commander told researchers: “There is nothing we haven’t received an answer on.”

According to the study, members of ISWAP and JAS were trained to use AI systems and bypass safeguards designed to prevent chatbots from responding to dangerous requests.

Former fighters said dedicated AI units were established within the groups, with members including engineers, weapons specialists and bomb-makers.

One former JAS commander described a 23-member specialist unit working from a solar-powered room, where members answered questions, analysed information and trained senior commanders on the use of AI.

A former Boko Haram commander said fighters consulted AI while making explosives, describing the technology as “like a human robot”.

Antonia Juelich, author of the Cambridge study, told Al Jazeera that about 10 of the 27 people interviewed discussed using AI in connection with explosives.

She said the concern was not only that AI could provide general weapons information, but that fighters could describe the specific materials and problems they faced and receive tailored responses.

The study also found that fighters reportedly consulted AI about recovering explosive material from unexploded military ordnance and adapting commercially available drones to carry explosives.

AI was also reportedly used during combat. In one case, a fighter transmitted images from the battlefield to commanders, who uploaded the footage to ChatGPT for analysis before giving instructions to those on the front line.

Former fighters also described consulting AI about unfamiliar weapons and tactics.

In one account, a commander asked an AI specialist how to operate newly acquired firearms and was reportedly told: “Just ask Grok.”

The technology was also reportedly used to assess failed operations. Fighters would describe their tactics to AI and ask why an attack had failed, while footage was uploaded for analysis to help determine how future operations could be conducted differently.

Separately, researchers from Tech Against Terrorism found evidence that supporters of the wider ISIL movement were discussing ways to bypass AI safety controls and obtain weapons-related information.

One user compared bypassing AI safeguards to removing a car’s speed limiter.

Lucas Webber of Tech Against Terrorism said researchers were seeing increasing discussion of attempts to “jailbreak” AI systems and warned that the material uncovered could represent only part of the activity.

Major AI companies have rules prohibiting the use of their systems for terrorism, violence and weapons development. Companies including OpenAI, Anthropic, Google, xAI, Meta and DeepSeek were contacted by Al Jazeera for comment on the findings.

The research comes amid growing concerns over terrorist groups’ use of emerging technologies. United Nations officials warned the Security Council in February about increased use of advanced AI by ISIL and its affiliates.

The Cambridge interviews largely covered AI use by former fighters during 2023 and 2024. Juelich said the rapid development of AI systems since then raised concerns about how the technology could be used by armed groups today.

“I worry about where this is heading,” she said.