Skip to content

R0043/2026-04-01/Q003/SRC03/E01

Research R0043 — Sycophancy Vocabulary
Run 2026-04-01
Query Q003
Source SRC03
Evidence SRC03-E01
Type Analytical

9-domain, 53-threat AI taxonomy that excludes sycophancy

URL: https://arxiv.org/html/2511.21901

Extract

The taxonomy organizes AI threats into 9 domains with 53 sub-threats:

  1. Misuse (prompt injection, jailbreaking, deepfakes)
  2. Poisoning (backdooring, label flipping)
  3. Privacy (model inversion, membership inference)
  4. Adversarial (evasion attacks)
  5. Biases (representational harm, allocational harm)
  6. Unreliable Outputs (hallucinations, factual errors)
  7. Drift (concept drift, data distribution shifts)
  8. Supply Chain (compromised models)
  9. IP Threat (model theft)

The taxonomy bridges technical and business language by mapping threats to business loss categories (Confidentiality, Integrity, Availability, Legal, Reputation) and aligning with NIST AI RMF, ISO/IEC 42001, and the EU AI Act.

Sycophancy is NOT included. The closest category is "Unreliable Outputs" (domain 6), but this covers factual errors (hallucinations), not agreement-seeking behavior. A sycophantic model producing factually correct but agreement-biased output would not be captured.

JUDGMENT: This taxonomy's omission of sycophancy is significant — it represents one of the most comprehensive cross-domain AI threat taxonomies available, yet it has a blind spot for behavioral model properties that prioritize agreement over accuracy. The gap persists even in efforts specifically designed to bridge domains.

Relevance to Hypotheses

Hypothesis Relationship Strength
H1 Partially supports Active taxonomy effort exists
H2 Contradicts Taxonomy efforts exist
H3 Strongly supports Even the most comprehensive taxonomy excludes sycophancy

Context

The 53-threat taxonomy's exclusion of sycophancy — despite including 53 other operationally defined threats — is strong evidence that sycophancy is not yet integrated into the cross-domain AI risk vocabulary, even by researchers specifically working on taxonomy bridging.