Explainable AI for the EU Right to Explanation: A Systematic Review of the Law-XAI Translation Gap
AI Summary
This systematic review examines whether explainable AI can operationalize the EU Right to Explanation under GDPR Article 15(1)(h), AI Act Article 86, and related rules. According to the supplied abstract, the authors screened 2,643 records, reviewed 57 full texts, and found only 19 papers with substantive legal-technical integration. They identify recurring errors concerning the GDPR legal basis, limited engagement with the CJEU’s Dun & Bradstreet judgment, and confusion between explanation form and content. The paper proposes an Addressee/Purpose Framework, a four-phase implementation blueprint, and six open research questions.
Why it's worth reading
As organizations prepare to operationalize AI Act obligations, the review maps concrete mismatches between legal purposes and current XAI techniques that could undermine meaningful compliance.
Deep Read
1. What happened
Original facts: According to the supplied abstract, this paper is a systematic review of explainable AI in relation to the EU Right to Explanation. It focuses on GDPR Article 15(1)(h), AI Act Article 86, and related instruments. The review covers papers published from 2024 onward, reflecting the July 2024 publication of the final AI Act and the late addition of Article 86.
2. Core technology and framework
Original facts: The authors introduce an “Addressee/Purpose Framework.” It separates explanation form, which depends on the intended recipient, from explanation content, which depends on the relevant legal purpose. They also propose a four-phase operationalization blueprint, although the abstract does not describe the individual phases.
Analysis: This framing treats legal interpretation as a design input for XAI systems: identify the rights-holder and legal objective before selecting what information to expose and how to communicate it.
3. Key evidence and numbers
Original facts: The authors report 2,643 initial records from a deliberately broad search, 57 full texts reviewed, and only 19 papers showing substantive integration of legal and technical perspectives. Three recurring problems are reported: incorrect identification of the GDPR legal basis, limited engagement with the CJEU’s Dun & Bradstreet judgment, and conflation of explanation form with explanation content.
Uncertainty: The abstract does not provide databases, search strings, exclusion criteria, inter-rater agreement, quality-assessment results, or the exact frequency behind terms such as “most.” Reproducibility therefore cannot be assessed from the supplied material.
4. Why it matters
Analysis: A legally meaningful explanation is not automatically equivalent to a feature-attribution chart, local surrogate model, or natural-language rationale. XAI outputs may be understandable yet still fail to help an affected person identify errors, challenge a decision, or seek an effective remedy.
5. Practical impact
Analysis: Engineering and compliance teams could use the proposed distinction to test whether an explanation’s audience, legal purpose, required content, and presentation are aligned. The review also points researchers toward joint legal-technical evaluation. The paper reportedly identifies six open questions, but the abstract does not enumerate them, so their contents should not be inferred.
6. Limitations and uncertainty
Original facts: The authors note that few reviewed papers engage with the Dun & Bradstreet judgment and suggest publication timing as a likely reason.
Analysis: Restricting the corpus to work published from 2024 onward improves relevance to Article 86 but may exclude earlier XAI research with continuing legal or technical value. In addition, the supplied arXiv identifier 2608.02699 and publication date of August 3, 2026 are future-dated relative to currently verifiable records. Until the full text is independently checked, the paper’s metadata, methodology, and conclusions remain unverified.