Hany Farid, a professor at the UC Berkeley School of Information and a leading expert in digital forensics, has stated that advancements in artificial intelligence image generation have made it increasingly difficult for experts to detect deepfakes. According to a report from The New York Times, Farid noted that AI has significantly improved in mimicking real-world elements, which has reduced the effectiveness of traditional detection methods.
Farid expressed his concerns about the rapid development of AI by major tech companies, suggesting that they prioritize profit over ethical considerations. He mentioned his frustration with the current state of technology and indicated a desire to retire from the field.
Data from cybersecurity firm DeepStrike indicates that the number of deepfake files has surged from 500,000 in 2023 to 8 million in 2025, coinciding with a significant increase in fraud attempts. In response to the challenges posed by deepfakes, experts are employing AI technologies to combat these threats, including tools like the Deepfake-O-Meter and initiatives like Content Credentials, which aim to establish standards for digital photo authenticity.
Additionally, emerging scams related to deepfakes include celebrity impersonations, catfishing, fake security footage, and manipulated political videos.