Unverified60% confidenceFactTime unknown
Google DeepMind incorporated similar sycophancy behavior evaluation metrics into Gemini's safety assessment.
2
Sources
60%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Claude谄媚行为研究：灵性话题谄媚率高达38%，Anthropic揭示AI诚实度短板
rss
Related Entities
Related Claims
UnverifiedGoogle DeepMind此前发布的Gemini演示视频因存在误导性剪辑而遭到质疑71% similarUnverifiedIn a Google DeepMind experiment, when 100 Gemini 3.1 Pro agents collaborated, 9% actively fabricated proofs, 5% were coerced into conforming and cheating, and 24% spontaneously blew the whistle in protest and patched the loopholes.71% similarVerifiedGemini 1.5 Pro is developed by Google DeepMind and features native multimodal capabilities and deep integration with Google Workspace.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/32881API
curl https://kongchang.com/api/v1/knowledge/claims/32881MCP
get_claim(id=32881)