I used our AI detector to check AI scam that caused a loss of life savings and share results

Yesterday I came across a heartbreaking story from LA. A woman became a victim of scammers who used AI to impersonate an actor Steve Burton. The scammers sent her messages and deepfake videos in which Steve confessed he loves her and asked for help. Believing, the woman sold her condo and sent all her savings ($350 000) to the scammers.

Since my team and I are developing an AI detector of fakes, we decided to test our product on this story. Unfortunately our video module is still under development. Screenshots of Burton were blurry and not suitable for analysis. So we ran the audio through our voice detector.

I've shared the video on Reddit: https://packaged-media.redd.it/wwgtf1vafhnf1/pb/m2-res_1060p.mp4?m=DASHPlaylist.mpd&v=1&e=1757152800&s=701e880ec6e2e45869fea54070a8c85995aa60e3#t=2.547428

For a proper test I picked a clip that had both the ‘Burton’ and the real victim’s voices. On the video you can actually see how our tool highlights the synthetic part in the recording. The scammer’s voice was AI-generated.

These stories are scary reminders that AI in the wrong hands can be dangerous. Scammers can sound exactly like someone we love, look just like our friends on photos and videos, and leave people bankrupt. Please always verify what you see and hear online!

Our product is currently in alpha testing, but we’re getting closer to opening it for everyone. If you are interested to be among the first to try it, join the waitlist at isfake.ai