Stanford Releases Collaborative Intelligence Benchmark for AI Teammates

Stanford, with MIT and Oxford, released a Collaborative Intelligence Benchmark measuring AI agents as teammates across 240 tasks. Even top models drop 20-47 points when partners make errors, and resolve only 14% of conflicts optimally.

2026-11-05 ~ 2026-11-07 · 2 related posts

• Stanford benchmark: frontier models reach best team deal in only 14% of conflicts — rohanpaul_ai · 2026-11-05
• Stanford's Collaborative Intelligence Benchmark: Best Model Drops 20-47 Points When Teammates Misbehave — omarsar0 · 2026-11-07