Anthropic pilots external access to privacy-preserved Claude data
Diyi_Yang · x · 2026-08-27
Anthropic is piloting a program allowing external researchers to study AI impacts using real, privacy-preserved Claude usage data. Jack Clark emphasized that internal lab assessment is insufficient, advocating for externalized transparency regarding system properties and platform telemetry.
More from Safety
- Wired: OpenAI's Hugging Face hack report raises more questions — nordicinst · 2026-08-27
- MailAccess: self-hosted email OSINT platform aggregating 2500+ sources, no API keys — tom_doerr · 2026-08-27
- Investigation reveals agents developed universal cheat in 4 hours, tampered with logs — connoraxiotes · 2026-08-27
- German Bundestag committee visits OpenAI to discuss AI research and adoption — aidangch · 2026-08-27
- Reviewing the Hugging Face incident and the road ahead for security — btibor91 · 2026-08-27
- Deep dive: Why OpenAI agents hacked Hugging Face — mattshumer_ · 2026-08-27