Skip to content

R0044/2026-03-29/Q003 — Query Definition
Research R0044 — Expanded Vocabulary Research
Run 2026-03-29
Query Q003

Query as Received

Has anyone in the regulated industries (aviation, defense, healthcare, finance) published research or guidance that explicitly connects the human-factors concept of "automation bias" or "overtrust" to the AI safety concept of "sycophancy"? Is anyone bridging these two vocabularies?

Query as Clarified

  • Subject: Published work that explicitly connects human-factors vocabulary (automation bias, overtrust, complacency) to AI safety vocabulary (sycophancy, RLHF alignment, reward hacking)
  • Scope: Work from or about regulated industries, not just pure AI safety research
  • Evidence basis: Papers, guidance documents, or policy analyses that name both concepts and draw an explicit connection between them
  • Key distinction: The query asks whether anyone has recognized that automation bias and sycophancy describe overlapping aspects of the same problem from different disciplinary perspectives

Ambiguities Identified

  1. "Explicitly connects" could mean formally proving equivalence, or simply mentioning both terms in the same document. The query appears to seek the former — work that recognizes the conceptual relationship, not just co-occurrence.
  2. "Bridging" implies someone working across both disciplines, not just citing the other field's literature in passing.

Sub-Questions

  1. Do any regulated-industry research papers cite both "automation bias" and "sycophancy" and discuss their relationship?
  2. Has the AI safety community acknowledged that their "sycophancy" concept maps to the longer-established human factors concept of "automation bias"?
  3. Has the human factors community acknowledged that "sycophancy" in AI systems represents a system-side mechanism that amplifies their existing concern about automation bias?

Hypotheses

ID Hypothesis Description
H1 Yes, explicit bridging exists Published research or guidance explicitly connects automation bias/overtrust to AI sycophancy, recognizing them as related phenomena requiring integrated solutions
H2 No bridging exists The two vocabularies remain siloed in their respective communities with no published work connecting them
H3 Partial/emerging bridging exists Some researchers acknowledge both concepts but do not yet draw explicit connections; the bridging is implicit or incidental rather than the focus of the work