Transparency on Paper, Black Box in Practice

Voice Assistants, the GDPR and the AI Act

Abstract

This work examines the European Union’s regulatory framework on transparency, focusing in particular on the dimension of explainability with specific reference to voice assistants. It analyzes the main legal sources governing this aspect, namely the General Data Protection Regulation and the Artificial Intelligence Act, with the aim of clarifying the scope and content of transparency and explainability obligations under both regulations. It investigates how these requirements operate in practice by applying the legal framework to a specific category of AI-based consumer devices: voice assistants. In this context, Amazon Alexa is used as a case study to determine the applicable legal regime and to assess how transparency obligations are translated into real-world interactions between users and AI systems. The analysis then considers Alexa+, the latest version of the technology, whose enhanced functionalities and increased autonomy raise some regulatory questions. In particular, the work explores whether its features could lead to classification as a high-risk AI system under the AI Act, triggering stricter obligations.