Probability of Being Targeted by a Deepfake

Deepfake technology has advanced rapidly, with deepfake incidents increasing by over 3,000% between 2022 and 2023. While comprehensive prevalence data is still emerging, surveys suggest that approximately 0.5% of American adults have been personally targeted by deepfake content, a number expected to grow significantly.

The most common malicious uses of deepfakes include non-consensual intimate imagery (accounting for about 98% of deepfake videos online, overwhelmingly targeting women), financial fraud (deepfake audio or video of executives used to authorize transfers), political disinformation, and identity theft. Deepfake-facilitated fraud losses exceeded $25 billion globally in 2023.

Detection of deepfakes is becoming increasingly difficult as the technology improves. Current detection methods include examining inconsistencies in blinking patterns, lighting, skin texture, and audio-visual synchronization, but these are becoming less reliable. Several states have enacted laws specifically targeting malicious deepfakes, particularly non-consensual intimate imagery. Prevention measures include watermarking original content, media literacy education, and emerging blockchain-based content authentication systems.