The Viral AI War Fakes: How Iran Used Deepfakes to Mislead the World
Following Iran’s missile strikes against Israel, a wave of AI-generated deepfake videos rapidly spread online. These fabricated clips, some allegedly created with Google’s Video3 model, depicted Israeli cities and infrastructure suffering immense damage, achieving virality despite their fictional nature.
Facts and Forensics
- The highly realistic fakes garnered millions of views on platforms like TikTok, X, and Instagram, with some reaching over 18 million views in a week.
- Forensic experts confirm the most widely shared videos are sophisticated deepfakes, often leveraging advanced AI video generation tools like Google’s Veo3, previously KaiFu Labs’ “Kling 2.1 Master,” and “Seedream.”
- These tools, including open-source options like “Wan 2.1,” allow the creation of videos from real images (Image-to-Video), producing “super-realistic” content capable of circumventing platform content restrictions, as noted by experts like Ben Buchanan.
The sophisticated AI-generated content represents a significant leap in deepfake quality, enabling widespread manipulation of public perception during the ongoing hostilities.
Campaign Tactics and Regional Divide
After the Israeli retaliatory strikes against Iran’s nuclear facilities, Tehran deployed a comprehensive AI misinformation campaign across multiple fronts within the first week:
Targeted Deceptions
- Transformations of peaceful Israeli neighborhoods into war zones.
- Showcases of Tel Aviv’s Ben Gurion Airport being struck and an El Al plane engulfed in flames.
- Humiliation narratives: Videos pitting Ayatollah Khamenei against Israeli PM Netanyahu and US President Trump, with AI creating realistic depictions of Khamenei symbolically dominating or humiliating the figures.
- Sympathy manipulation: Depicting Iranian families mourning alongside high-profile missile tests and footage used to position Iran’s top cleric as a symbol of strength.
- Propaganda Tool Co-option: Utilizing fabricated historical footage (e.g., Chile wildfires passed off as Israeli cities burning) through official channels like Iranian TV.
- Facilitation through Escapes: AI-generated content proliferating rapidly on channels and accounts operating outside the purview or with the cooperation of major social media platforms.
Pun intended. These horrifically edited visuals are racking up massive view counts. Easily doubling the real attack footage views. I was just on an X livestream… it’s unbelievably sophisticated.
The International Institute for Counter-Terrorism (ICT) in Israel observed strategic language use: Arabic/Farsi content emphasizes solidarity against Israel, while Hebrew-centric narratives aim for psychological impact within Israel, demonstrating a localized approach to global disinformation.
Moving Beyond Static Lies: The Rise of Intelligent Deception
This conflict showcases the evolution beyond simple image manipulation or news satire into complex generative AI warfare.
The Generative AI Arms Race
Advanced capabilities, including sophisticated text generation (“pro-Russia articles”) and video editing tools, allow perpetrators to create hyper-realistic narratives tailored to specific audiences, bypassing traditional detection methods.
The evidence is clear and
terrifying. These AI systems are being weaponized on an unprecedented scale. Beyond the Middle East, global actors are exploiting these tools to sow discord and manipulate narratives.
The problem is compounded by lucrative markets; KBV Research predicts the virtual influencer market could reach $37.8 billion by 2030, highlighting the versatility—and danger—of these technologies.
Conclusion: The Battle for Truth Intensifies
In this information war, AI-generated disinformation poses an existential threat by weaponizing the very tools used for communication. Widespread misuse erodes trust and hinders effective conflict resolution. Detecting malign AI manipulation in real-time remains a critical challenge for society.
