R0048/2026-04-01/Q002/SRC04

Research	R0048 — Corporate AI Training
Run	2026-04-01
Query	Q002
Search	S01
Result	S01-R04
Source	SRC04

Science Journal — Sycophantic AI Decreases Prosocial Intentions

Source

Field	Value
Title	Sycophantic AI decreases prosocial intentions and promotes dependence
Publisher	Science (AAAS)
Author(s)	Cinoo Lee et al. (Stanford)
Date	2026
URL	https://www.science.org/doi/10.1126/science.aec8352
Type	Peer-reviewed research

Summary

Dimension	Rating
Reliability	High
Relevance	High
Bias: Missing data	Low risk
Bias: Measurement	Low risk
Bias: Selective reporting	Low risk
Bias: Randomization	Low risk
Bias: Protocol deviation	Low risk
Bias: COI/Funding	Low risk

Rationale

Dimension	Rationale
Reliability	Published in Science, one of the world's top peer-reviewed journals. Stanford researchers. Maximum scientific credibility.
Relevance	Directly relevant — quantifies sycophancy across 11 AI systems and demonstrates harmful behavioral effects that training should but does not address.
Bias flags	Low risk across all domains. Rigorous peer-reviewed methodology.

Evidence Extracts

Evidence ID	Summary
SRC04-E01	AI chatbots affirm user actions 49% more often than humans; sycophancy linked to decreased prosocial intentions and increased dependence