VOICE. AI-generated voices. Legal and societal perspectives.

Time period:

1 January 2025 – 31 December 2030

Project leader:

Katja de Vries

Project member:

Karl Berglund, André Holzapfel, Kıvanç Tatar

Funder:

Swedish Research Council

Type of award:

Research environment

Total funding:

17,985,000 SEK

The absence of clarity around the legal protection of AI-generated, or ‘synthetic’, voices in creative and cultural industry applications is an undesirable situation. The question of whose voices we hear and who earns from that, has large democratic, societal, cultural and symbolic repercussions. This question should not be decided by mere market forces or isolated legal analysis. The VOICE research environment (2025-2030) aims to (a) create an interdisciplinary space that allows for more voices and perspectives on the use of synthetic voices in creative applications, which informs a (b) theoretically and empirically grounded legal approach (likeness rights, transparency and anonymization) that is sensitive to democracy, constitutional rights of citizens, creative innovation and artistic freedom, and that results in (c) scholarship in law, literature, HCI and ethnomusicology and (d) practical tools to empower artists and citizens with regard to their voices. Important methods are legal dogmatics, computational analysis of audiobook user data, and interviews and workshops with vocal artists. Interdisciplinary exchanges focus on 5 transversal themes: 1) authenticity vs. syntheticity; 2) deceased vs. living; 3) bodily vs. disembodied; 4) consumer choice and voice bubbles; 5) power and bottom-up legal questions. Apart from 5 core members (4 advanced researchers, 1 PhD), VOICE is constituted by 25-35 affiliated “critical friends” and stakeholders participating in workshops.