DF26 benchmark: humans and detectors near random on AI speech videos

Researchers released DF26, a deepfake detection benchmark of 271 real and 2,420 AI-generated speech videos. Both humans and top detectors performed near chance, exposing fragility against modern generative models.

2026-09-10 ~ 2026-09-11 · 3 related posts

• DF26 benchmark: humans and detectors near chance on AI speaking videos — Severyn Shykula · 2026-09-10
• DF26 benchmark shows humans and top deepfake detectors are near random at spotting AI videos — ducha_aiki · 2026-09-10

1 near-duplicate retellings: Justgototheeffinmoon