Deepfakes becoming most of recorded AI fraud

Statement (NASA form)

Given that synthetic media already outnumbers AV, face-ID, and moderation incidents combined, there is a possibility of fraud desks still staffed for last decade’s attack mix resulting in losses and civic injury that the control book still treats as a side channel.

Public tallies now put deepfakes as the majority of recorded AI-fraud cases. The residual moved; many mitigations did not.

Composite 16 = 4×3 + 4