AI-Generated Deepfake of Missile Attack Tests Limits of Digital Forensics

On a Sunday morning, Hany Farid received an email containing a link to a video that appeared to show a U.S.-made missile striking an elementary school in Iran, allegedly killing more than 150 people, most of them children. The clip had already amassed at least 1.1 million views on social media. Farid opened the file, noting realistic camera shake, accurate shadows, and a plausible delay between the missile impact and the sound of an explosion.

New York Times

Farid’s recent work has documented a surge in AI-generated videos depicting bombings, plane crashes, fires and executions. His research indicates that most people can no longer reliably differentiate authentic photographs, audio recordings or video clips from synthetic creations. The proliferation of generative AI has lowered the technical barrier to producing convincing visual misinformation.

New York Times

For more than twenty years, Farid has been a leading expert in digital forensics, assisting governments, human-rights organizations, journalists, law-enforcement agencies and thousands of private individuals. He routinely evaluates requests to authenticate visual media, a workload that has intensified as AI tools become more accessible.

AI Expert Futurist Keynote Speaker: LLMs ML

• The viral video recorded a missile traveling at roughly 500 mph, with a brief pause before the explosion sound, matching the speed of sound.

New York Times

• Farid reported reviewing “dozens” of similarly convincing AI-generated videos in the preceding days.

New York Times

• The email accompanying the clip included the notes: “Is this an internet hoax or an international war crime?” and “We’re trying to verify what’s real.”

New York Times

Farid has warned that AI is increasingly obscuring factual reality, distorting perception, and threatening democratic discourse. He described his own experience as a gradual loss of visual certainty, stating that the technology is “fracturing democracies” and that he feels his ability to discern truth is deteriorating.

New York Times

• “Is this an internet hoax or an international war crime?” — Email note accompanying the video

New York Times

• “We’re trying to verify what’s real.” — Email note accompanying the video

New York Times

• “Anyone can create a video of anything or anybody, doing or saying anything,” — Hany Farid, digital forensics expert

New York Times

• “This will take a little time.” — Hany Farid, responding to the verification request

New York Times

• “I feel like I’m going blind,” — Hany Farid, describing his struggle with AI-generated media

New York Times

The incident illustrates how synthetic media can rapidly achieve viral reach, potentially shaping public opinion before verification is possible. When visual evidence can be fabricated at scale, the capacity of institutions and citizens to assess credibility is undermined, raising concerns about misinformation’s impact on policy decisions and public trust.

New York Times

Facing mounting personal and professional strain, Farid and his wife are planning to leave California for a farm in rural Vermont. He intends to continue his forensic work remotely, emphasizing the need for improved detection tools and public awareness as AI-generated content proliferates.

AI Expert Futurist Keynote Speaker: LLMs ML