Hany Farid, a leading expert in digital forensics and deepfake detection, is grappling with a growing crisis of trust in the authenticity of visual content amid the rapid rise of artificial intelligence-generated media. Based in Berkeley, California, Farid has spent more than two decades specializing in identifying manipulated images and videos, aiding governments, law enforcement, human rights groups, and journalists in distinguishing fact from fabrication. However, in recent months, he has found himself increasingly uncertain—even about his own expert analyses.
The turning point came with a viral video purportedly showing a U.S.-made missile striking an elementary school in Mina, Iran, killing over 150 people, mostly children. Farid meticulously examined the footage, assessing the physics, shadow geometry, and sound delay, and geolocating the scene using vast online image databases. Despite the intricate scrutiny, which included calculating missile size and trajectory, Farid could not definitively declare the video real or fake. His team found no compelling evidence of alteration, yet the emergence of numerous AI-generated false videos on related events complicated the picture.
Farid’s skepticism reflects a broader challenge: AI technologies have become capable of fabricating highly convincing videos, images, and audio clip, blurring the line between reality and illusion. These synthetic media have been proliferated not only by malicious actors abroad but also, he says, through official channels in the U.S., heightening concerns about misinformation during sensitive political and social moments. Farid himself was recently targeted when his voice was cloned and used in fraudulent calls, highlighting the personal risks associated with the technology.
His wife, Emily Cooper, a vision scientist at the University of California, Berkeley, shares the unease, observing that the crisis extends beyond digital content. She points to the increased screen time and decreased exposure to natural environments contributing to a surge in eye health problems, including myopia and potential blindness—a growing public health concern linked to modern lifestyle changes.
Faced with the accelerating pace of AI-driven deception and its consequences, Farid and Cooper are planning to relocate from California to a rural property near Dartmouth College in New Hampshire. There, they hope to find a healthier environment for themselves and their work. While Farid remains involved in research and consulting, the shift underscores his growing disillusionment with Silicon Valley’s role in fueling technologies that undermine trust and societal cohesion.
Despite his concerns, Farid continues to teach and engage with students, emphasizing the urgent need to understand the technology’s potential and limitations. He warns that while deepfake creation is now easy and inexpensive, detection remains costly and complex, leaving society vulnerable. When asked if there was a near-term solution to the problem, he admitted that while solutions may exist, the situation is dire.
“We’re pretty screwed,” Farid said, reflecting on the vast challenges posed by AI-generated misinformation and the implications for truth in the digital age.