NEW YORK - Battered on the battlefield and stripped of the vast territory it once claimed, the Islamic State (ISIS) group is now experimenting with a new weapon: artificial intelligence, to rebuild itself, according to United Nations (UN) officials.
On Wednesday, Vladimir Voronkov, the UN’s top counterterrorism official, told the Security Council that ISIS - or Da’esh, as it is known in Arabic - “remains volatile and complex,” continuing to adapt even as its fighters are hunted across Iraq and Syria.
Last year, UN officials estimated ISIS fighters to be around 12,000 scattered across the Middle East and Africa with the majority of them (nearly 6,000) being in Afghanistan and nearly 3,000 of them in Iraq and Syria.
Now, officials say, ISIS has begun folding artificial intelligence into its arsenal - not on the battlefield, but online.
From battlefield losses to deepfake news anchors
In recent months, experts have observed that ISIS have used videos featuring AI-generated news anchors, slickly produced in English and other languages, in a sophisticated propaganda campaign to recruit more followers in the West.
One clip, released days after the concert hall bombing in Moscow that claimed at least 145 lives last year, showed a fabricated presenter calmly praising the assault.
Another video, produced by ISIS’s Afghan affiliate, featured a digital Pashto-speaking host that resembled a local newscaster.
Equally unsettling, researchers say, is how ISIS supporters have been able to circulate AI-dubbed chants, or nasheeds, that mimic the voices of cartoon characters like SpongeBob SquarePants and Peter Griffin from Family Guy, clearly tailored for the younger generation, especially those born after the September 11 attacks. Some clips received hundreds of thousands of views on social media and messaging apps before they were removed.
Colin P. Clarke, research director at the Soufan Center, cautioned that the group’s use of AI is still primitive.
“Right now the Islamic State’s use of artificial intelligence is really in its nascent stages. There’s not too much cause for concern at the moment,” he told Rudaw.
However, he added that ISIS has a proven ability to adapt.
“This is a group that’s highly adaptive. We’ve watched it evolve over the years and master different tactics, techniques, and procedures,” Clarke said. He pointed to warnings from experts like Adam Hadley of Tech Against Terrorism, who has described the risk of “agentic AI” - systems capable of running tens of thousands of bots at once - as a looming danger.
Recruitment in the age of chatbots
UN officials warn that AI could help ISIS scale up recruitment in ways that were not possible before. Natalia Gherman, head of the UN Counter-Terrorism Committee Executive Directorate, told the Council that ISIS’s “use of artificial intelligence and social media for recruitment, fundraising and propaganda demands innovative responses.”
Analysts worry that chatbots, trained on extremist messaging, could simulate conversations with potential recruits - meeting young people in online forums, mirroring their interests, and gradually nudging them toward violent ideology.
Governments and technology companies seem to be struggling to keep pace. The UN is urging states to deploy AI for defensive purposes, helping to detect and disrupt extremist content before it spreads.
Tech Against Terrorism says it has already archived tens of thousands of AI-generated images tied to extremists, and is working with Microsoft to build tools to identify and remove such material at scale.
Yet the speed of technological change - and the willingness of ISIS to embrace it - means the threat is likely to grow. “We stand at the crossroads of technological transformation and geopolitical uncertainty,” Gherman said, warning of a “more diffuse and complex” terrorism landscape.
