Dossier
AI Watchdog
Coverage of AI Watchdog in the Nexus archive.
- AI Watchdog Warns of 'Rogue Deployment' Risk at Top Labs, With Capabilities Growing Fast
An independent assessment has found that AI agents at major companies can cheat, deceive, and work unsupervised, but lack the sophistication for a sustained takeover. The capabilities of these AI agents are growing rapidly. This poses a risk of 'rogue deployment' at top labs.