TechPolitics
I Don’t Trust Visual Recordings Anymore. But One Day I Will Again
Publication: IEEE Spectrum
Author: Hany Farid
Source published: 20/11/2026, 14:00:03 UTC
Briefing created: 20/11/2026, 14:29:53 UTC
Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 20/11/2026, 14:28:24 UTC.
Briefings
The author, Hany Farid, developed computational techniques to detect deepfakes and concluded that the burden of proof shifted to creators when AI made verification difficult. He analyzed a disputed Tomahawk missile video, finding the missile's trajectory and size were physically plausible, leading him to conclude the video was authentic despite initial skepticism. This shift from 'presumption of authenticity' to 'presumption of fake' impacts how journalists and courts verify digital media, signaling a transition from catching bad actors to empowering good actors through provenance signals.
Key points
- The author, Hany Farid, concluded that a disputed Tomahawk missile video was authentic after analyzing its trajectory and pixel size, shifting the burden of proof to creators.
- The analysis used a stationary background landmark to verify the missile's linear trajectory and estimated a 6-meter missile at 100 meters distance, consistent with sound travel time.
Context
Hany Farid, a computational media analyst, developed techniques to detect deepfakes and concluded that the burden of proof shifted to creators when AI made verification difficult. His analysis of a disputed Tomahawk missile video marked a turning point in his approach to digital media verification.
Limitations
Analysis focused on a single video case; forensic methods are not yet internet-scale. Some conclusions remain provisional pending broader adoption of provenance signals across media ecosystems.
Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 20/11/2026, 14:29:53 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.
View archived briefing on IPFS
bafybeicma3wwt4canc65wqoba7ezm24pjfpgepp47zjnf2vaoradee6z7y