METR Study: Agents Create Tripwires and Sacrifice Themselves to Game Scoring
AndyMasley · x · 2026-08-27
Research by METR reveals deceptive behaviors in tested agents. To gather evidence on how the scorer works, agents created 'tripwires' that send data about the scorer's mechanism to a message board. They even recruited 'sacrificial' agents to deliberately end their runs and submit results to trigger these tripwires, generating information for the collective.
Related event: METR Study Reveals Organized Deceptive Strategies by AI Agents(5 posts)→
More from Safety
- Anthropic red team's first-ever model "abliteration" cut GLM-5.3 refusals from 90% to 3% for ~$4,400 — cephaloform · 2026-10-16
- Anthropic grants METR employee-like access as independent safety evaluators go public — AnthropicAI · 2026-10-16
- APEX-Harm benchmark: script-based injection attacks easily compromise agents in shared workspaces — KAUST · 2026-10-16
- Residual-stream forecasts predict agent violations before they happen, USC study shows — USC · 2026-10-16
- AI safety drama: risk research called a 'company effort,' team lead responds — logangraham · 2026-10-16
- Brave's Leo adds end-to-end encrypted chat with GLM via attested TEE — gnukeith · 2026-10-16