Nigeria: More than 20 former members of the jihadist movement Boko Haram have revealed how the group is utilizing advanced artificial intelligence technologies to bolster their operations, as reported in a recent study by Cambridge University. The research, led by Antonia Juelich, a specialist in technology and terrorism, provides a rare insight into the tactical use of AI by these groups, emphasizing the evolving strategies of jihadist factions in Nigeria.

According to France24.com, the report indicates that Boko Haram has been integrating AI technologies since early 2023, shortly after the launch of AI models like ChatGPT. The group has developed specialized units dedicated to AI, using these tools not only for propaganda but for enhancing their tactical capabilities. A former jihadist disclosed to Juelich that AI is perceived as an all-knowing entity, providing guidance on using new weapons, improving explosives, and planning raids.

The study highlights a training regime where instructors from regions like Iraq and North Africa teach members to effectively engage with Large Language Models (LLMs), leveraging techniques such as "jailbreaking" to bypass ethical constraints embedded in AI systems. This allows jihadists to extract sensitive information that could be applied to their militant activities.

Graig Klein, an expert in international terrorism at Leiden University, corroborates these findings, suggesting that AI acts as a "force multiplier" for jihadist groups, enhancing the efficiency of their operations rather than creating new forms of terrorism. Klein warns of the real threat posed by AI in the hands of terrorist organizations, as it can make their strategies more effective and difficult to counter.

The report also underscores the inadequacy of current AI safeguards in preventing misuse by terrorist groups. Klein explains that with persistence and specific prompting techniques, these groups can circumvent existing protective measures. Both Juelich and Klein stress the urgency of reinforcing AI safeguards to prevent the potential development of weapons of mass destruction by such groups.

The findings reflect a broader concern that the use of AI by terrorist organizations is not confined to Nigeria but is likely part of a global strategy by groups like ISIS to enhance their operational capabilities. This underscores the pressing need for the international community to address the vulnerabilities in AI technologies to prevent their exploitation for violent ends.