The AI Act proposal: a new right to technical interpretability?

The EU is developing a legal framework for AI explainability, but gaps remain: The paper examines whether the proposed AI Act, combined with existing GDPR and Convention 108 plus provisions, adequately establish a "right to technical interpretability" for EU citizens—suggesting current legislation may be insufficient to ensure people can understand how AI systems make decisions affecting them.

Current legal protections focus on explanation rights rather than technical understanding: The existing GDPR Article 22 addresses a "right to explanation," but the paper indicates this may not translate into meaningful technical explainability, highlighting a distinction between being told a decision was made by AI versus understanding the technical reasons why.

Multiple legal instruments are being layered together without clear coordination: The paper notes that the AI Act proposal, GDPR, and Convention 108 plus operate in combination, but it's unclear whether this patchwork approach adequately protects citizens' ability to understand AI systems or if more explicit legislation is needed.

The gap between legal rights and technical reality needs addressing: The research bridges legal scholarship (focused on GDPR) and technical scholarship (focused on explainable AI techniques), suggesting that neither field alone has solved the problem of ensuring genuine technical interpretability for affected individuals.

EU policymakers should consider explicitly codifying technical explainability rights: The paper's central question—whether the EU should include a specific right to technical explainability in current legislation—suggests this protection is not yet guaranteed and warrants deliberate policy action.