Skip to content

R0043/2026-03-28/Q003

Query: Has the vocabulary gap itself been identified as a problem in the AI safety or AI governance literature? Are there researchers or organizations working to create a shared taxonomy that bridges AI safety terminology with regulated-industry terminology?

BLUF: The broader AI terminology gap is well-recognized, with multiple active bridging efforts (MIT AI Risk Repository with 1,600 risk formulations, AIR 2024 with 314 risk categories, Standardized Threat Taxonomy with 53 threat vectors). However, the specific sycophancy/overreliance vocabulary gap between AI safety and regulated industries has not been articulated as a distinct problem. Sycophancy is absent from every major bridging taxonomy examined.

Answer: H3 (Gap recognized but not for sycophancy) · Confidence: Medium


Summary

Entity Description
Query Definition Question as received, clarified, ambiguities, sub-questions
Assessment Full analytical product
ACH Matrix Evidence × hypotheses diagnosticity analysis
Self-Audit ROBIS-adapted 4-domain process audit

Hypotheses

ID Statement Status
H1 Gap recognized and actively addressed Partially supported (general gap only)
H2 Gap not recognized Eliminated
H3 Gap recognized but not for sycophancy Supported

Active Bridging Efforts

Organization Effort Scale Sycophancy Coverage
MIT FutureTech AI Risk Repository 1,600 risk formulations / 65 documents Not explicitly named
AIR 2024 researchers Risk Categorization Decoded 314 categories / 24 policies "Overreliance" noted as underspecified
arXiv researchers Standardized Threat Taxonomy 53 threats / 9 domains Explicitly absent
Trilateral Research AI Terminology Gap article Practical 6-step solution Not mentioned
arXiv researchers Vocabulary neologisms paper Theoretical framework Argues all existing terms are insufficient

Searches

ID Target Type Outcome
S01 AI taxonomy bridging efforts WebSearch 5 selected / 15 rejected
S02 Sycophancy-specific vocabulary gap WebSearch 0 selected / 10 rejected — null result is a finding

Sources

Source Description Reliability Relevance Evidence
SRC01 Trilateral terminology gap Medium High 1 extract
SRC02 Standardized Threat Taxonomy Medium-High High 1 extract
SRC03 MIT AI Risk Repository Medium-High High 1 extract
SRC04 Vocabulary neologisms Medium High 1 extract
SRC05 AIR 2024 Medium-High High 1 extract

Revisit Triggers

  • Publication of a taxonomy that explicitly includes sycophancy as a cross-domain risk category
  • OECD or ISO release of AI behavioral characteristics vocabulary
  • Professional society (IEEE, ACM) standardization of sycophancy-adjacent terminology
  • New bridging research that explicitly connects "sycophancy" with "automation bias" as the same phenomenon