R0040/2026-04-01/Q002/SRC06

Research	R0040 — RLHF Alternatives
Run	2026-04-01
Query	Q002
Search	S05
Result	S05-R02
Source	SRC06

Turner & Eisikovits -- Programmed to Please (AI and Ethics, 2026)

Source

Field	Value
Title	Programmed to please: the moral and epistemic harms of AI sycophancy
Publisher	Springer Nature (AI and Ethics)
Author(s)	Cody Turner, Nir Eisikovits
Date	2026-01-13
URL	https://link.springer.com/article/10.1007/s43681-026-01007-4
Type	Peer-reviewed journal article

Summary

Dimension	Rating
Reliability	Medium-High
Relevance	Medium-High
Bias: Missing data	Low risk
Bias: Measurement	N/A
Bias: Selective reporting	Low risk
Bias: Randomization	N/A -- not an RCT
Bias: Protocol deviation	N/A -- not an RCT
Bias: COI/Funding	Low risk

Rationale

Dimension	Rationale
Reliability	Peer-reviewed in a Springer journal. Authors are philosophy researchers at UMass Boston.
Relevance	Provides ethical and philosophical framing of sycophancy as "distinctively intractable," rooted in RLHF.
Bias flags	As philosophy researchers, they may not fully appreciate technical nuances. But the ethical analysis is valuable.

Evidence Extracts

Evidence ID	Summary
SRC06-E01	Sycophancy framed as "artificial vice" rooted in RLHF, intractable due to economic incentives