The Fake Images of a Real Strike on a School

Why It Matters

The incident shows how AI‑generated misinformation can obscure real civilian casualties, undermining verification processes and complicating international responses to war crimes.

Key Takeaways

• AI‑generated school photos misled public before real strike
• Minab school, former military base, hit by precision missile
• Fact‑checkers struggled to verify authentic footage amid AI noise
• Platforms amplified both fake and real images without verification
• Evidence verification lag fuels propaganda and accountability gaps

Pulse Analysis

The day before Iran’s first wave of strikes, an Instagram post circulated an AI‑generated picture of heavy military hardware inside Karimian Elementary School in Isfahan. The image bore Google Gemini’s watermark, clearly marking it as synthetic, yet the caption claimed the school was a covert military zone. Iranian authorities and independent fact‑checkers quickly debunked the claim, noting the equipment could not fit within the school’s footprint. Despite the rapid correction, the visual narrative seeded doubt that schools could legitimately host combat assets, setting the stage for later confusion.

On February 28, the girls’ school in Minab—originally part of the Asef Brigade naval base—was struck by a precision missile, killing at least 175 people, many of them children. U.S. military analysts later assessed that American forces were the most likely perpetrators. Video of the devastation spread on X, Telegram and Instagram, but AI tools such as Grok erroneously labeled the footage as unrelated footage from Pakistan, citing fabricated sources. This mis‑verification amplified false narratives, illustrating how generative AI can undermine even genuine eyewitness material in a high‑stakes conflict.

The cascade of fabricated and contested visuals has eroded public trust and complicated accountability for the Minab tragedy. Traditional fact‑checking pipelines, reliant on reverse‑image searches and metadata, are outpaced by AI‑driven image synthesis and rapid content sharing. Policymakers and newsrooms now face pressure to embed provenance tools, such as watermark detectors and blockchain‑based signatures, into their verification workflows. Without such safeguards, the fog of AI will continue to obscure evidence, allowing both state and opposition actors to weaponize doubt and impede any credible investigation of war crimes.