Deepfake Discovery: How to Find Your Face

Deepfake intimate images cannot be completely avoided, any publicly accessible face photograph is a potential source for deepfake generation. However, proactive steps help reduce both the likelihood of being targeted and the harm if an attack occurs. These steps include controlling the public availability of high-quality face photographs, implementing technical protections that make deepfakes harder to create convincingly, and monitoring for any content that appears online using your likeness.

Reducing your deepfake exposure

Steps to find your face in deepfakes include: setting social media profiles to private; avoiding high-resolution face photographs in public posts; limiting face photographs in professional directories to low-resolution versions; removing yourself from people-search databases; and avoiding video call screenshots being taken by others. None of these prevent all risk, but they reduce the quality of material available to perpetrators.

Technical discovery tools

Emerging technical discovery tools include: photo cloaking tools (Fawkes, Glaze) that add adversarial noise to images making them resistant to AI training; digital watermarking tools that embed traceable identifiers in your photos; C2PA (Coalition for Content Provenance and Authenticity) content credentials that certify authentic images; and screen recording protection on smartphones.

Ongoing monitoring

ScanErase's monitoring service conducts regular scans across 2.4 billion indexed images to find any content using your likeness before it spreads widely. Early detection, finding content within days of its first appearance rather than months later, dramatically reduces the harm. The 14-month average discovery time for NCII victims means most damage occurs during the unmonitored window.