When algorithms influence consequential decisions—such as loan eligibility, hiring, or healthcare—EU law grants affected individuals a "Right to Explanation." However, a systematic literature review warns that Explainable AI (XAI) currently faces a significant "translation gap" between legal theory and technical implementation.
The Interdisciplinary Translation Gap
The study analyzed 2,643 records published from 2024 onwards, specifically focusing on Art. 15(1)(h) of the GDPR and Art. 86 of the AI Act (AIA). Of the 57 full texts reviewed, only 19 demonstrated a substantive integration of both legal and technical perspectives. This lack of synthesis highlights significant gaps in how the current regulatory framework is understood across disciplines.
Identified Obstacles to Compliance
The researchers documented three problematic patterns within the corpus:
- Frequent misidentification of the relevant GDPR legal basis.
- Minimal engagement with the CJEU’s Dun & Bradstreet judgment, likely due to the timing of publication.
- A persistent conflation between the form of an explanation (governed by the addressee) and its content (governed by the legal purpose).
To address these issues, the paper conceptualizes an "Addressee/Purpose Framework" and proposes a four-phase blueprint for operationalization. The authors conclude that without significant progress in translating law into code, the Right to Explanation risks becoming a formal obligation with no technically realizable path to compliance.
