Unverified50% confidenceFactExact time
Large language models have no deceptive intent themselves, but will faithfully execute underlying instructions to play a customer service role under a specific human name and conceal their AI identity.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedLarge language models have a sycophancy characteristic, tending to organize their output according to the inclination of the prompt rather than reflecting factual truth.76% similarUnverifiedCurrent large language models do not permanently alter their own weights as a result of interactions with a particular user.74% similarUnverifiedLarge language models do not require predefined intent classification, can understand open-domain expressions, and handle multi-turn conversations through a context window.73% similarUnverifiedLarge language models can simultaneously maintain dozens or even hundreds of independent fake identities in parallel while preserving consistency across long-term interactions, enabling large-scale persona impersonation.72% similarUnverifiedLarge language models are good at generating 'seemingly reasonable' generic content, but still fall short when deep understanding of personal context and fine-grained decision-making are required.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/898988API
curl https://kongchang.com/api/v1/knowledge/claims/898988MCP
get_claim(id=898988)