Claim Verification: "Deepfake videos are now indistinguishable from real footage to the average human eye." — Disproved

Automated fact-verification of the claim: "Deepfake videos are now indistinguishable from real footage to the average human eye." Verdict: DISPROVED Key Findings Average humans detect deepfake videos above chance level, with accuracy ranging from 57% to 67% across major studies — well above the 50% threshold that "indistinguishable" would require. A large-scale study (N=1,901) found human video discrimination AUC of 0.67, with humans outperforming AI algorithms on video deepfakes specifically. A separate UK-based study (N=1,093) confirmed that "people are better than random at determining whether an individual video is genuine or fake." Leading deepfake researchers explicitly distinguish between voice deepfakes (which have crossed the "indistinguishable threshold") and video deepfakes (which have not). Files proof.py — Re-runnable Python verification script proof.md — Structured proof report proof_audit.md — Full verification audit trail proof_narrative.md — Plain-language summary proof.json — Machine-readable structured data Generated by Proof Engine v1.2.0.