I discussed with João Pedro Quintais the European Commission's Guidelines on transparency obligations for providers and deployers of certain AI systems and the application of Article 50 of the EU AI Act more broadly. João also commented on how copyright considerations and history can make us aware of challenges that may arise in the context of the AI Act.
To find the European Commission's Guidelines, see here: https://digital-strategy.ec.europa.eu/en/policies/guidelines-ai-transparency-obligations.
Throughout the episode, João commented on some of the complementary guidance that AI providers can also take into account when implementing watermarking. To access them directly, click on the links below:
- 01:55, Code of Practice on Transparency of AI-generated Content: https://digital-strategy.ec.europa.eu/en/policies/code-practice-ai-generated-content.
- 06:11, The General-Purpose AI Code of Practice: https://digital-strategy.ec.europa.eu/en/policies/contents-code-gpai.
- 06:11, The Guidelines for providers of general-purpose AI models: https://digital-strategy.ec.europa.eu/en/policies/guidelines-gpai-providers.
- 11:23, Research indicating that identification of material as AI-generated can reduce its perceived accuracy: https://arxiv.org/html/2506.16202v2.
- 12:27, Anthropic's approach to watermarking: https://www.anthropic.com/news/claude-text-watermark.
Key moments
00:00 Introduction to EU AI transparency rules
00:25 João Pedro Quintais' background and research focus
01:07 Scope and application of Article 50 of the AI Act
01:55 The necessity of Guidelines and their legal space
02:48 Difference between Guidelines and the Code of Practice
04:25 Hierarchy of AI regulation: Law, Guidelines, and Codes
05:41 Institutional influence on digital regulation
06:24 Operational details in the Guidelines and ecosystem logic
07:54 Legal and technical governance fusion in AI regulation
08:58 Transparency, labeling, and trust in AI content
10:00 Provenance versus trustworthiness in AI labeling
11:14 Impact of labeling on perceptions of AI-generated content
12:05 Familiarity with AI and perception shifts13:14 Implementation of watermarking in AI systems
15:00 Robustness and detection challenges in watermarking
16:10 Technical and normative challenges in AI marking
17:28 Copyright law, authorship, and AI outputs
19:01 AI labeling as a legal and evidentiary signal
21:06 Influence of copyright and legal language in AI regulation
22:15 Closing remarks and future discussions in AI law
To learn more about Dr. Alba Ribera Martínez and her research, we invite you to visit her website: https://www.albariberamartinez.com.
She also writes, from time to time, in her newsletter, the DMA Agora, about the latest developments surrounding the European regulation: https://www.linkedin.com/build-relation/newsletter-follow?entityUrn=7344021393451184128.
________
The Binary Agora is on social media, where we upload and comment on dedicated content only for our subscribers. Stay tuned there, as well.
- LinkedIn: https://www.linkedin.com/company/thebinaryagora/posts/?feedView=all- BlueSky: https://bsky.app/profile/thebinaryagora.bsky.social
- TikTok: https://www.tiktok.com/@thebinaryagora
- Instagram: https://www.instagram.com/thebinaryagora/
The views and opinions expressed by our guests are their own and do not necessarily reflect the official position of this podcast, its host, or any affiliated organisations.