ADF STAFF
Terrorist groups are experimenting with AI to spread their ideology and make attacks more deadly. But security professionals are finding that the tool is just as valuable when used to protect citizens from these groups.
“AI-powered surveillance tools can track terrorist movements, analyze suspicious financial transactions and detect radicalization patterns on social media platforms,” Mahmoud Ali Youssouf, African Union Commission chairperson, said during a 2025 gathering of the AU Peace and Security Council.
Al-Qaida and the Islamic State group have embraced online technology, and each has sought to create a “digital caliphate” that reaches across international borders and far beyond anything a physical caliphate would achieve.
Widely used AI systems such as ChatGPT, Grok and Claude have become key pieces of the digital caliphate strategy, helping terrorists plan attacks, edit video footage and translate propaganda into different languages to make it available across ethnic groups and countries.
AI reduces the time needed to produce propaganda, and that AI-produced material can, in turn, radicalize small groups or individuals who might never have been radicalized. Blockchain technology helps terror organizations encrypt their messaging to evade security agencies.
“The potential for tech-savvy militants to build unrestricted jihadi chatbots may make violent knowledge available on demand, providing a virtual planner that guides lone actors through every step of an attack,” researchers Danielle Cosgrove, Doug Livermore, Erin K. McFee, Morgan Tadych and Timothy “Tito” Torres wrote recently for the Atlantic Council.
Even as terrorists use AI for their purposes, Africa’s governments and security agencies have the capacity to employ their own AI technology to identify and undermine terrorism. Studies have found that some of the same AI systems, known as large language models (LLM), that help terrorists also can be turned against them to detect online extremism.
According to the United Nations Interregional Crime and Justice Research Institute (UNICRI) and the United Nations Counter-Terrorism Centre (UNCCT), governments can use AI-based technology to counter terrorist messaging and activities including: predicting terrorism, identifying radicalization risks, detecting terrorist-generated misinformation and disinformation, automating content moderation to identify and remove propaganda, countering terrorist claims, and analyzing extensive data sets.
These methods become more important as AI accelerates terror groups’ capacity for sowing discord.
“As the timeline between exposure to radical content and committing acts of violence shortens, efforts must be aimed at limiting access to the content in the first place,” Cosgrove and others wrote.
One method of disrupting terrorist content is known as “hash sharing.” The technology tags harmful content with a unique number that can be shared among law enforcement agencies and online companies to track and remove it.
“Hash sharing has been proven effective in preventing the exploitation of children online,” Cosgrove and others wrote. “Applying the same concepts to counterterrorism, and using existing counterterrorism partnerships between organizations and countries, could represent a pathway to quickly identify and remove content before it spreads widely.”
So far, security agencies worldwide remain behind the AI curve. A survey by UNICRI and Interpol found that half of the 50 law enforcement agencies polled reported that their AI expertise was “rudimentary.” Only 20% reported their capacity as “advanced.”