Hello!

We are saying goodbye to the Polish Council Presidency and welcoming the Danish. We expect they will make child protection a centrepiece of their work on digital files. Meanwhile, both the Parliament and the Commission are trying to alleviate the tension between copyright and AI, in a non-legislative way for now.

Dimi, Michele and Iglika https://meta.wikimedia.org/wiki/User:Pelajanela (who is visiting us for the week in Brussels)

=== Age Verification ===

EU Age Verification: As per our latest count a total of 14 EU countries have now expressed support for a “digital age of majority” and mandatory age verification for accessing social media platforms. The group, however, seems split along a very important fault line - the age limit. France, Greece and Denmark want it to be at 15. Spain wants 16. Austria wants to leave it up to the individual countries, while Belgium and Slovakia want it harmonised.

—

Who should do it? As regular readers will know by now, the other major debate is who should be obliged for age verification. The first part of this question is who should perform the age verification: service providers want app stores and operating systems to be in charge, app store providers want services to be on the line. Google and Meta are having a very public brawl https://blog.google/around-the-globe/google-europe/age-assurance-europe/ over this. The second part of this question is which services should be covered. One side wants only the highest risk services to be covered (adult content and gambling), the other wants social media to be included as well. There isn’t a political consensus on either part yet.

—

Italian Age Verification: At the national level, Italy is pushing to introduce age verification https://www.senato.it/show-doc?leg=19&tipodoc=DDLPRES&id=1418922&idoggetto=0&part=ddlpres_ddlpres1 measures and a minimum age of 15 for the use of social media. The project is still a draft, but Wikimedia projects would end up in scope if passed as is. Wikimedia Europe and Wikimedia Italia are coordinating on informing lawmakers about this, the consequences and the possible solutions. We remind you that a while ago the French Loi SREN had a similar provision, but thanks to the work of Wikimédia France https://www.wikimedia.fr/le-projet-de-loi-reguler-et-securiser-lespace-numerique-une-nouvelle-contrainte-pour-un-internet-libre-et-ouverte/ the definition was changed to exclude Wikipedia.

—

Why it matters for Wikimedia: Mandatory age checks for public interest, general knowledge projects like ours would constitute an additional barrier to access to knowledge. It would also mean that more user data needs to be collected, one way or another.

=== CSAM Fallback ===

The CSAM Regulation, the one mandating the scanning, reporting and removal of child sexual abuse material, is solidly stuck in Council. Countries can’t agree how to handle messaging services. Despite strong ambitions, it does not look likely that the upcoming Danish Council Presidency will be able to change this. Platforms currently make use of an exemption in the form of a derogation to the e-Privacy Directive, as a stop-gap measure to legally scan for CSA materials. But this derogation is set to expire in April 2026. France, Ireland and Hungary are now proposing to extend it.

—

Why it matters for Wikimedia: The Wikimedia Foundation takes CSAM seriously and has procedures to mitigate against it. So do many of our communities. A change in the legal framework could require us to adjust our own actions. Beyond that, it is an important topic for user safety and for the safety of our projects.

=== Danish Council Presidency ===

The Danish Presidency of the Council https://danish-presidency.consilium.europa.eu/ is kicking off tomorrow and their digital priorities are squarely on protecting children online. We expect the Danish to consolidate and lead the group of 14 countries that demand age verification for social media into some sort of initiative. When it comes to CSAM, as explained above, they will be forced to work on the stopgap extension if they don’t manage to flip the position of at least a major country (e.g. Germany) on the issue of scanning direct messages.

—-

The Danish government also wants to work out a well-coordinated EU approach to multilateral negotiations, notably on the open, secure and free internet and the implementation of the Global Digital Compact, something our movement has been actively involved https://meta.wikimedia.org/wiki/Global_Digital_Compact_Wikimedia_Advocacy_Collaboration_2024 with.

—-

The Presidency will facilitate a discussion on the updating of the Audiovisual Media Services Directive (something we are following, as for instance, not all rules that apply to Netflix should also apply to Wikimedia Commons). They are also organising a conference on artificial intelligence and copyright in September in Copenhagen.

=== AI & Copyright ===

The European Parliament is working on the thorny issue of generative AI and copyright in a non-legislative report. German conservative lawmaker Axel Voss expected to publish his first draft in mid July. The final version is expected in December. The file might go a long way in forming some sort of political consensus, if that is at all possible. The one major question to be answered is whether lawmakers see a need to re-open the EU’s copyright rules. The European Commission doesn’t seem eager to touch this file this legislative term, but Renate Nikolay (Deputy Director of DG CNECT https://commission.europa.eu/about/departments-and-executive-agencies/communications-networks-content-and-technology_en) said that “there might be a need to revisit the copyright rules” during a meeting with lawmakers in the civil liberties committee.

—

We are expecting the AI Act’s code of practice for general-purpose AI models to be finalised this week https://digital-strategy.ec.europa.eu/en/policies/ai-code-practice. The code of practice is supposed to be an additional agreed upon set of high-level commitments and detailed measures to implement them. It will have commitments on transparency and copyright. It will be interesting to see if the rights reservation system under the EU’s text and data mining exception is laid out in more detail.

—

Why it matters for Wikimedia:

Copyright laws define to a significant extent what content we can share on our projects. Generative AI models have become a disruptive part of the knowledge ecosystem. They use our projects and are used by our editors and readers. We are part of this development and discussion, whether we like it or not.

=== Danish Proposal to Extend Copyright to Faces ===

The Danish government is proposing https://kum.dk/fileadmin/_kum/1_Nyheder_og_presse/2025/Aftale.pdf to extend copyright protection to everyone’s own body, facial features and voice. The idea, so the claim goes, is to put citizens in a better legal position to fight against deep fakes.

—

According to Danish Wikimedian Matthias Smed Larsen “the purpose of the proposal is to give it a statutory basis with the purpose of making it easier to utilize takedown obligations under EU law. In my opinion putting it in the Copyright Act is a misnomer.”

—

WMEU is reaching out to WMDK to see what could or should be done. In any case, the draft legislation is not published yet and we are waiting to see the actual text.

=== Wikipedia Test ===

The Global Advocacy Team at the Wikimedia Foundation has published The Wikipedia Test https://medium.com/wikimedia-policy/the-wikipedia-test-75b41bcc8583. A public policy tool to help ensure regulators consider how new laws affect online communities and platforms that provide services and information in the public interest. It can be used by Wikimedians when communicating with lawmakers.

===END===