Deepfake expert Hany Farid warns AI will soon cripple visual truth

Hany Farid, a pioneering researcher in digital forensics, has spent more than 20 years developing tools to detect deepfake images and videos. Speaking to the New York Times, he warned that advances in artificial intelligence now allow anyone to create convincing false media, overwhelming the capacity of experts to verify authenticity. Farid, who collaborates with governments, human‑rights groups, journalists and courts, says the volume of fabricated material has grown from a few cases per week to countless daily uploads, forcing analysts to spend much longer on each item. He predicts that within one to two years the “entire visual system” of the public could become unreliable. Detection now requires checking lighting, sound, shadows, subtle skin‑color changes, and even blood‑flow patterns. Farid himself fell victim to AI‑generated voice impersonation used in an attempt to obtain confidential information, underscoring the personal risk posed by the technology.