Skip to content

R0021/2026-03-25/Q007

Query: What published research exists on AI decision auditing, explainability requirements, or judgment logging in automated AI systems?

BLUF: Substantial research exists. DARPA invested in a 4-year XAI program (2017-2021). The EU mandates AI explainability through GDPR and the AI Act. Over 2,400 XAI papers were published between 2022-2025 alone. However, practical deployment faces challenges — post-hoc explanation methods are approximations, and traceability constraints hinder regulatory compliance.

Answer: H3 (Active, growing, but challenges remain) · Confidence: High


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 Substantial research with practical frameworks Partially supported
H2 Research is minimal or theoretical Eliminated
H3 Active and growing but challenges remain Supported

Key Programs and Frameworks

Program Scope Status
DARPA XAI Explainable AI research (2017-2021) Completed; XAITK toolkit released
GDPR Article 22 Right to explanation for automated decisions Active law
EU AI Act Article 86 "Clear and meaningful explanations" requirement Effective 2024-2025
XAI Research Field 2,425 papers published 2022-2025 Rapidly growing

Searches

ID Target Type Outcome
S01 XAI research papers WebSearch 3 selected, 7 rejected
S02 DARPA XAI program WebSearch 3 selected, 7 rejected

Sources

Source Description Reliability Relevance Evidence
SRC01 DARPA XAI Program High High 1 extract
SRC02 EU AI Act/GDPR High High 1 extract
SRC03 XAI Research Review High High 1 extract