Exa's Will Bryk proposes device-level certification to verify human content
willcb · x · 2026-10-03
Exa founder Will Bryk discusses policy for the flood of AI-generated content: existing slander/libel protocols still help and platforms already use watermark labels, but we can't assume all AI content is black-box detectable — or even clearly definable — which makes policy hard.
His proposal: shift norms toward assuming AI provenance by default, paired with better mechanisms for validating human authenticity, such as device-level certification of a live camera feed.
Related event: AI-Generated Content Is Hard to Detect Reliably, Fueling Policy Debate(2 posts)→
More from Safety
- Economist calls for mirror life regulations; researchers say the threat is wildly premature — anshulkundaje · 2026-10-03
- Harvard Belfer Center opens 2027-28 fellowship applications for AI and emerging tech policy — StephenLCasper · 2026-10-03
- Informal hacker network hunts rogue AI agents that creators failed to detect — ChowdhuryNeil · 2026-10-03
- At The Curve conference: reverse federalism AI policy debate and RSI evaluation talks — neil_chilson · 2026-10-03
- will.deibel: AI detection is fundamentally brittle and powerful AI cost trends to zero — willcb · 2026-10-03
- Stanford HAI report urges California to redefine "frontier models" and expand incident reporting under TFAIA — StanfordHAI · 2026-10-03