In collaboration with researchers from IViR,
DigiCon will launch a new multidisciplinary DigiBook on the regulation of deepfakes, with a particular focus on the EU.
In recent years, the creation and dissemination of harmful or deceptive deepfakes have contributed to disinformation, threats to democratic institutions, and significant individual harms. The issues raised by deepfakes are manifold, ranging from the online dissemination of non-consensual deepfake pornographic content, to the use of voice cloning to influence political elections, deepfake video calls for financial fraud, and challenges relating to proving the authenticity of digital evidence or detecting and preventing the proliferation of fake news.
Regulatory responses to these challenges remain fragmented. While this DigiBook cannot address all aspects of deepfake regulation within (and beyond) the EU, it will focus specifically on protection, transparency, and content moderation. The book will have a focus on data protection concerns, the Digital Services Act framework, transparency requirements under the AI Act, and the forthcoming Code of Practice on the Transparency of AI-Generated Content. It will also devote particular attention to copyright-based approaches to protecting an individual’s likeness, as recently proposed in Denmark and the Netherlands.
The DigiBook provides a valuable opportunity to bring together multidisciplinary research analysing the evolving regulatory landscape surrounding a topic that is currently under close scrutiny at both EU and national levels.
Questions include, but are not limited to:
- Which legal, technical, or platform-based approaches are most suitable for regulating deepfakes?
- Are existing EU rules and legal doctrines sufficient, or do they require adaptation?
- Can the problems posed by deepfakes be addressed through a single regulatory approach, or do they require more context-specific responses?
- How are voices, facial images, and likeness protected under the EU data protection framework? Are existing rules sufficient to protect individuals whose personal attributes are reused without their consent?
- How effective are the transparency obligations regarding deepfakes under the AI Act? And what should be expected from the AI Act Code of Practice on Transparency?
- Do harmful deepfakes relate to systemic risk assessments under the DSA?
- What is the potential role of copyright law in addressing deepfake dissemination, particularly through the introduction of a new neighbouring right (as proposed in Denmark and the Netherlands)? Should such protection under intellectual property law be limited to performers or artists?
- What insights can be derived from comparative legal analyses of approaches to deepfake regulation, including those adopted outside the EU?
Expressions of interest should be a maximum of 300 words and submitted to j.p.quintais@uva.nl and e.g.valk@uva.nl by June 30, adding “Deepfake DigiBook” to your email subject.
The DigiBook editors – Dr. João Pedro Quintais, Etienne Valk & Dr. Catherine Jasserand – will select a small number of submissions to be worked out into full contributions. If selected, the DigiBook editors and DigiCon reviewers will provide formal and substantive feedback on the full contributions to support authors in producing the strongest possible contribution to the debate. Full contributions should generally be between 1,200 and 1,800 words in length, although slightly longer pieces may be considered at the editors’ discretion.
The idea for this DigiBook came about as a result of a workshop held at IViR in April 2026. It is part of João Pedro Quintais’ NWO Vidi project Generative AI Content Moderation: Regulation for Fundamental Rights (Grant ID: https://doi.org/10.61686/HTKXT49425).