The AI dilemma facing IS and al-Qaeda - The Global Jigsaw podcast, BBC World Service

Key Concepts

• Artificial Intelligence (AI): The development of computer systems that can perform tasks typically requiring human intelligence, such as learning, problem-solving, and decision-making.
• Jihadist Media: Propaganda and communication materials produced by extremist groups aiming to promote their ideology and recruit followers.
• IS (Islamic State) and al-Qaeda: Prominent jihadist organizations.
• ISKP (Islamic State Khorasan Province): An affiliate of the Islamic State group based in Afghanistan.
• AI Hallucinations: Instances where AI generates inaccurate or fabricated information.
• Fard al-'ain: An Islamic religious term meaning an individual obligation on every able Muslim, often used in grave circumstances like defense.
• Deepfakes: Synthetic media where a person in an existing image or video is replaced with someone else's likeness, often used for misinformation.
• Cognitive Warfare: The use of information and psychological operations to influence the thinking and decision-making of adversaries.
• Media Jihad: The use of media and propaganda for jihadist purposes.
• Haram: Prohibited in Islam.

AI and Jihadist Propaganda: An Evolving Landscape

This summary details the evolving use of Artificial Intelligence (AI) by jihadist groups, particularly the Islamic State (IS) and al-Qaeda, as explored by BBC Monitoring's Jihadist Media Team. The discussion highlights both the potential benefits and significant reservations these groups have regarding AI, alongside the broader concerns raised by AI experts like Sam Altman and Yoshua Bengio about the technology's potential for misuse.

Concerns about AI's Potential for Harm

Sam Altman, creator of ChatGPT, and Yoshua Bengio, a prominent AI figure, express grave concerns about the rapid advancement of AI. Altman states, "I’m very concerned that we are accelerating on a road that we've never been. We could lose control very easily. My worst fear is that we, the industry, cause significant harm to the world. I think, if this technology goes wrong, it can go quite wrong." He elaborates on potential dangers, including AI providing weapons to terrorists, creating new pandemics, and destabilizing geopolitical stability, emphasizing that "We are not equipped to deal with this."

Jihadist Adoption of Technology

Historically, IS supporters have been early adopters of technology, with many, especially younger ones, being tech-savvy. Sympathizers of IS and al-Qaeda have been actively exploring AI, with publications encouraging followers to learn about it. They view AI as a "real boon" for spreading their messaging to younger and wider audiences.

Current AI Applications in Jihadist Media

While AI use by jihadists has been "inconsistent" and "patchy" so far, they are actively experimenting and learning. Current applications primarily include:

• Translations: AI is frequently used for translating IS and al-Qaeda materials, particularly IS's weekly newspaper Al-Naba, into English and other languages to reach a global, mainly English-speaking audience. This aims to make complex, classical Arabic content more accessible.
• Content Enhancement: AI is used to enhance official texts from IS or al-Qaeda by pairing them with visuals and audio to make them more appealing, especially to younger audiences.
• Poster and Video Repackaging: AI assists in repackaging existing content, including overlaying commentary on archive IS footage.
• Transcription: Videos are quickly transcribed into detailed blow-by-blow accounts with narration.
• Social Media Enhancement: AI is used to enhance and spread IS messaging on platforms like Instagram and Facebook, often indicated by an "AI tick" on posts.

Experimental AI-Generated Content

More advanced experiments include:

• Animated News Bulletins: A pro-IS media outlet, "Hisad," produced animated news bulletins featuring an AI-generated newsreader. This outlet shared reports and narratives from Al-Naba.
• AI-Generated Newsreader: The newsreader in Hisad's bulletins was not a real person but AI-generated.

Ideological and Practical Reservations Among Jihadists

Despite the potential, jihadists exhibit considerable caution regarding AI, particularly in cognitive warfare. Several factors contribute to these reservations:

• Aesthetics and Taste: Some IS supporters found the AI-generated news bulletins "puerile" and akin to "YouTube shorts," arguing that their message is grave and should not be trivialized with "cheap cartoons and gimmicks."
• Religious Prohibitions (Haram):
  • The AI newsreader's uncovered face was considered haram in strict Islam.
  • The AI-generated voice, potentially based on a non-believer, was problematic for reading holy scripture.
  • The clothing of the AI character was criticized for not being Islamic.
• Accuracy and Hallucinations: A significant concern is the accuracy of AI translations. Jihadists worry about AI "hallucinating" or mistranslating Quranic verses and religious scriptures, which could be "catastrophic" and backfire. This concern is shared by media organizations, emphasizing the need for human oversight.
• Authority and Fatwas: Jihadist groups are alarmed by the idea of supporters using AI tools like ChatGPT to seek religious rulings (fatwas), as they prefer followers to consult accredited clerics within their own structures.
• Reputation and Credibility: The risk of AI backfiring and damaging their image and reputation is a major concern. The use of fake blood in a 2015 IS video from Yemen, which was exposed by al-Qaeda and led to ridicule (dubbed the "Vimto Caliphate"), serves as a cautionary tale about maintaining credibility.
• Guardrails in AI Tools: AI platforms like ChatGPT and Copilot have built-in guardrails that prevent them from generating content related to incitement, bomb-making, or violent acts, which can limit the production of violent footage.

ISKP's Stance on AI

ISKP (Islamic State Khorasan Province), based in Afghanistan, through its affiliated media group al-Azaim and its publication Voice of Khorasan, has shown a more explicit engagement with AI. The June 2025 edition featured an 11-page article on AI with the slogan: "learning about AI today is fard al-'ain." This designation, typically reserved for grave religious obligations, suggests a serious consideration of AI adoption. However, this specific phrasing was later removed due to backlash from rivals who mocked the group for applying such a significant religious term to technology. Despite this, it indicates that ISKP views learning about and adopting AI tools as significant.

Differences Between IS and al-Qaeda

• IS: Generally seen as more enthusiastic about AI, with more instances of its use by supporters. They are perceived as enjoying "gory propaganda" and are less flexible in their messaging.
• al-Qaeda: Considered a more "academic" and "mature" group, al-Qaeda is more reserved. They are not against AI but are still exploring the best ways to use it to present a compelling message. They tend to avoid excessively gory images.

AI as a Weapon vs. Propaganda Tool

Jihadists are more focused on how AI can be weaponized (e.g., in drones, surveillance) rather than solely for media use. They are keen to learn how to leverage AI for future military applications, though they are likely to do so discreetly.

The Role of Social Media Platforms

Jihadists face a dilemma between secure platforms and the need for reach. While they have been pushed off major platforms like Twitter, they are now experimenting on platforms like TikTok and Instagram, which are popular with younger audiences and focus on visuals and short videos. This presents a challenge for platforms like Meta and TikTok, which have policies against extremist content.

The "Ruslan" Case Study

An IS supporter named "Ruslan" emerged with striking AI-made material, offering his services to create slick, game-like videos. His productions, though amateurish, appealed to younger audiences interested in gaming. These videos featured animated IS militants in global landmarks with slogans. Ruslan's account disappeared, possibly due to increased online security warnings and paranoia, or potential arrest.

Misinformation and Deepfakes

The potential of AI and deepfakes for misinformation campaigns is a significant concern. The creation of seemingly authentic but fake images or videos (e.g., depicting the burning of the Quran or insults to Islam) can be used to incite attacks. An example cited is an AI-generated image of Druze fighters firing on a mosque, which jihadists shared to fuel sectarian violence, despite it being debunked.

Conclusion and Future Outlook

Jihadist groups are currently in a phase of learning, experimenting, and cautiously adopting AI. While they are excited by its potential, their use is tempered by ideological reservations, concerns about accuracy, and the need to maintain their image and credibility. The built-in guardrails of AI tools and platform policies also present challenges. It is highly likely that they will exploit AI further, particularly for weaponization and discreetly enhancing their capabilities. The future may see a more integrated use of AI in their propaganda, especially on visual-centric platforms like TikTok and Instagram, as they navigate the complexities of this powerful technology. The ISKP's emphasis on "responsible use" and avoiding disinformation, even while framing AI learning as a religious obligation, highlights the internal debates and strategic considerations surrounding AI adoption.