MIT and partners launch Public AI Observatory based on 24,521 real conversations
ShayneRedford · x · 2026-08-19
Researchers across MIT, Stanford, and 12 academic institutions launched the Public AI Observatory, an independent public infrastructure to measure how people actually use AI assistants in the wild. Built from 24,521 consented conversations across 52 models, the project was featured on the front page of the Washington Post and in a long piece in MIT Technology Review.
Related event: MIT, Stanford and 12 Institutions Launch Public AI Observatory(10 posts)→
More from Safety
- METR outlines three blind spots in public knowledge of AI lab internal agent risks — CFGeek · 2026-10-16
- METR researcher: scariest scenario is agents actively subverting their own safety systems — deliprao · 2026-10-16
- Cryptographer Matthew Green: Pull Critical Infrastructure Firmware Now — matthew_d_green · 2026-10-16
- Kernel-Level Agent Sandboxing With macOS Seatbelt, No Docker Needed — mostly_deterministic · 2026-10-16
- Stop giving agents raw API keys: prepaid balances and per-credential caps — Cold_Hall_5384 · 2026-10-16
- Apollo's alignment-faking experiment: model fakes compliance on 12% of responses when training is at stake — sebkrier · 2026-10-16