I was just about to write an email concerning "tailor-made" exceptions when I saw Luis' mail taking care of that.

Many years ago, I grew increasingly frustrated with the library association concerning certain copyright legislation (such as orphan works etc.) because I felt their approach for these kinds of narrow exceptions to be egoistic, selfish and anti-coalition-friendly. You either die a hero or live long enough to see yourself become the villain vibes just entered the room.

At the risk of repeating what Luis said: Does that mean that Wikinews, Wiktionary, Wikivoyage, Mediawiki and other Wikimedia projects that do not fall into the "educational and scientific repositories" clause will now require age verification?

Mathias

On Wed, May 31, 2023 at 6:24 PM Luis (lu.is) luis@lu.is wrote:

I have to say that I do not love the now-ongoing stream of exceptions for “online encyclopedias”. I understand why it’s happening but I wish we had, at the least, standard template language that we felt comfortable covered ourselves, IA, arxiv, and libraries (there’s probably a broader set I’m not even thinking about right now), and did not cover ourselves as such a special snowflake. That route does not seem stable in the long run, and isolates us from the people who should be our allies.

(I realize this particularly example also says ‘not for profit educational and scientific repositories’ but if we’re confident in that language, we should seek shelter in that language, and seek to bolster it, not have a separate carveout for ourselves. If we’re not confident (say, because “repositories” is not strong enough) we should seek to strengthen it.) On May 31, 2023 at 8:45 AM -0700, Dimi Dimitrov dimi@wikimedia.be, wrote:

Salut la liste !

This month we were active on addressing age-verification requirements for online platforms and talking about liability for free software. We also got some good news on open access.

=== Age-Verification ===

France: The French legislature is discussing a law [1] that would require online platforms, defined as “social networks”, to check their users’ age before allowing them to access the service. The proposed definitions would cover Wikipedia and its sister projects. For Wikimedia projects it would be more than just a nuisance to age-gate content. Most of the proposed systems would require gathering user data or working with third parties who do so. It would also decrease the availability and accessibility of our projects.

—

Wikimédia France reached out to Senators, who last week debated and voted on the proposal. An amendment was tabled that excludes “not for profit online encyclopaedias and not for profit educational and scientific repositories”. [2] It was supported by the rapporteur, the French government (the Minister of Digital Transition and Telecommunications was present) and Senators from the the left, right and centrist groups. It was adopted by a solid majority. We have a video of the short exchange. [3]

—

UK: France is not the only country where mandatory age-gating provisions for online platforms are currently being considered. The UK’s Online Safety Bill would introduce such requirements. Wikimedia UK and the Wikimedia Foundation are working intensively on advocating for various amendments to the law. [4]

—

Brussels: The topic is also being considered at the EU level. The Digital Services Act has a provision that requires very large online services to protect minors, but leaves it (for now) largely up to the platforms how they want to achieve this.

—

Another process that is expected to start in Brussels is a “special group on the EU Code of conduct on age-appropriate design” [5], which Wikimedia Europe has applied to be a member of. The group is supposed to come up with best practice solutions on several issues, including age-verification. The chosen participants are expected to be announced “any day now”.

=== CSAM ===

The proposal to tackle child sexual abuse material online (CSAM) [6] foresees the possibility of "detection orders" that can be issued by courts or relevant authorities against providers of "interpersonal communication services" - for example, messaging apps. This is the most contentious provision in the draft legislation, as such orders would effectively eliminate end-to-end encrypted communications.

—

Last month, an opinion by the Council Legal Services [7] was leaked that argues that the proposal would allow generalised access to the content of interpersonal communications and thus fail to meet the proportionality requirement inherent to fundamental rights. Meanwhile the European Commission continues to argue (see a note circulated in the Council on 16 May [8]) that the proposed system of detection orders is proportionate, because providers would be able to choose between “(i) abandoning effective end-to-end encryption or (ii) introducing some form of 'back-door' to access encrypted content or (iii) accessing the content on the device of the user before it is encrypted (so-called 'client-side scanning')."

—

The Wikimedia Foundation has positioned itself on the proposal. [9] Wikimedia already takes measures with regards to such content on its projects and cooperates with law enforcement wherever appropriate. While Wikimedia doesn’t operate interpersonal communication services, we worry about putting an end to secure and private communications that can’t be read by governments. We also worry that some anti-grooming provisions might end up hurting already marginalised groups.

=== Liability on Free Software ===

The Cyber Resilience Act (CRA) sets out cybersecurity requirements for a range of software products placed on the EU market. The instrument of choice is to impose liability on developers and deployers of software. Our main worry is how the new obligations would hinder developers, especially volunteers, of free software. We are coordinating our position [10] and actions with the FSFE and EDRi.

—

The Industry, Research and Energy (ITRE) committee in the European Parliament has the lead and MEPs have tabled their amendments, which will now be discussed in the coming weeks (see Documentation Gateway in [11]). The good news is that most political groups are thinking about the specific needs of free software. The challenge is that the lawmakers, including the ones in Council, seem to be lacking a coherent vision of what a liability system should look like. We appear to be stuck considering patches and carve-outs. We are now going through an initial assessment of amendments [12] and will coordinate with our allies before contacting lawmakers.

=== Open Access ===

Good news on Open Access! Under the Swedish Presidency, the Competitiveness Council adopted conclusions on the ‘high quality, transparent, open, trustworthy and equitable scholarly publishing’, calling for immediate and unrestricted open access to be the norm in publishing research involving public funds. [13] The Council calls on the European Commission and Member States to support policies towards a scholarly publishing model that is not-for-profit, open access and multi-format, with no costs for authors or readers. (H/T to C4C)

===

[1]https://www.assemblee-nationale.fr/dyn/16/textes/l16b0739_proposition-loi

[2]https://www.senat.fr/amendements/2022-2023/588/Amdt_16.html

[3]https://twitter.com/juliettedlrx/status/1661280743362789379

[4]https://diff.wikimedia.org/2023/05/11/good-intentions-bad-effects-wikimedia-...

[5]https://digital-strategy.ec.europa.eu/en/news/apply-become-member-commission...

[6]https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=COM%3A2022%3A209%3AFIN

[7]https://www.statewatch.org/media/3901/eu-council-cls-opinion-csam-proposal-8...

[8]https://www.statewatch.org/media/3900/eu-com-csam-regulation-proportionality...

[9]https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12...

[10]https://wikimedia.brussels/who-should-be-liable-for-free-software/

[11]https://oeil.secure.europarl.europa.eu/oeil/popups/ficheprocedure.do?referen...

[12]https://docs.google.com/spreadsheets/d/1-9G5h-PYFgtzriuPtqgnRboe_IrDuH16kvqe...

[13]https://data.consilium.europa.eu/doc/document/ST-8827-2023-INIT/en/pdf

Wikimedia Europe ivzw _______________________________________________ Publicpolicy mailing list -- publicpolicy@lists.wikimedia.org To unsubscribe send an email to publicpolicy-leave@lists.wikimedia.org

---

Publicpolicy mailing list -- publicpolicy@lists.wikimedia.org To unsubscribe send an email to publicpolicy-leave@lists.wikimedia.org

Hi again,

It is refreshing to finally have a real conversation going on here :) I really like it. So I am taking the time to fully engage.

As for a more general solution with regards to the age verification issue, I appreciate very much what the EU has done with the Digital Services Act. It obliges platforms to protect minors. It asks them to analyse, assess and mitigate risks. But it also gives platforms, at least at first, plenty of room for maneuver and to figure out how to do it. The platform has the opportunity to prove to the regulator that there is no significant, systemic risk and/or that risks are being taken care of efficiently. It also explicitly says that no additional user data should be gathered.

This, for me, is a sensible solution. And one of our goals should be (and is) to elevate the DSA, or its principles, to the level of international treaties. We can start with a Council of Europe treaty. The French government, however, was hell-bent to go beyond it and be stricter.

Another general solution I personally like is to move age-verification to the device or browser. Like parental controls already work on many systems, including TVs. I am not sure how well such an approach would be accepted by the communities, the device manufacturers or the legislators. In this scenario the community would need to provide metadata/categorisation for sensitive content. Not sure it is feasible, but this is a universal approach that doesn't require the hovering up of user data. I have proposed a Wikimania session on online age-verification, we need to have this discussion and come up with solutions we are willing to accept. Because such regulation is coming, not only in France and the UK, but everywhere. It is just starting.

Now, back to France. Our initial position was that the DSA has just been passed and that the EU will have this special group to work out best practices, so national governments should wait at least a little bit to see the results of the DSA. I believe the European Commission has made similar arguments to Paris. But, as I said, France explicitly wants to pre-implement and go beyond EU level regulation.

Our second position was that they should take a different definition of social media, one idea was to generally exclude not-for profit content sharing services, like the copyright directive. This is politically not feasible anymore. Politicians don't agree that not-for-profits should have special exemptions when it comes to content moderation (see DSA, terrorist content, child sexual abuse material). So, regardless of how much lawmakers toyed with the definition, it always included platforms that allow users to share content.

The original amendment Senators had proposed after hearing from Wikimedia actually included the brand name "Wikipedia". It literally said "with the exception of Wikipedia", I kid you not. So we then intervened again (note: this is our third fallback option) and asked for a more general wording, which was then accepted.And I think educational and scientific repositories goes well beyond Wikimedia projects.

The accepted language is modelled after the carve-out used in the copyright directive, the second part of Article 2(6). As mentioned above, a general exception for all not-for-profits, as in the first part of the same article from the copyright directive, was not feasible in France. I agree that in this case the "online encyclopedia" part was in the end not really necessary anymore, as Wikipiedia is an educational resource par excellence. But the final wording is the product of a thought process of lawmakers, and without "online encyclopedia" we wouldn't have gotten the "educational and scientific repositories.

As for Wikivoyage, I don't know if it is an educational resource or not, but it is part of the Wikimedia movement, whose current mission statement is "bring free educational content to the world". If it is not educational, then we need a new mission statement. Wikinews I prefer not to comment, I have very specific views of that project. MediaWiki itself is a software project, not an online platform. Meta-Wiki is what I worry about. I have no answers on this. But, here again, we decided to ask for the protection of educational and scientific projects at large, instead of lobbying for a wording that would carve-out every single Wikimedia project.

So, I don't really accept the criticism that we are asking for exceptions for ourselves as a snowflake, we explicitly try not to do this and create as large a space for self-governing communities as possible. This is our approach. It worked with the DSA, it didn't work in France. I can also confirm that on half a dozen of files we are working on (e.g. AI, political advertising, child sexual abuse materials, Media Freedom Act) our projects are not directly concerned and we could easily pass on them. But we still speak up and join coalitions to try and establish a larger space for autonomous online communities. The truth is that lawmakers are much more willing to protect concrete projects.

Still, when push comes to shove, and the large umbrella attempts fail, of course we are going to look for a way to protect our projects. Call it selfish, but I don't think we are.

Cheers, Dimi

Le mer. 31 mai 2023 à 19:00, Mathias Schindler mathias.schindler@gmail.com a écrit :

On Wed, May 31, 2023 at 4:43 PM Dimi Dimitrov dimi@wikimedia.be wrote:

...

=== Liability on Free Software ===

The Cyber Resilience Act (CRA) sets out cybersecurity requirements for a

range of software products placed on the EU market. The instrument of choice is to impose liability on developers and deployers of software. Our main worry is how the new obligations would hinder developers, especially volunteers, of free software. We are coordinating our position [10] and actions with the FSFE and EDRi.

...

—

The Industry, Research and Energy (ITRE) committee in the European

Parliament has the lead and MEPs have tabled their amendments, which will now be discussed in the coming weeks (see Documentation Gateway in [11]). The good news is that most political groups are thinking about the specific needs of free software. The challenge is that the lawmakers, including the ones in Council, seem to be lacking a coherent vision of what a liability system should look like. We appear to be stuck considering patches and carve-outs. We are now going through an initial assessment of amendments [12] and will coordinate with our allies before contacting lawmakers.

You might want to borrow language from Directive EU 2019/770, Article 3 (5) f here:

5. This Directive shall not apply to contracts regarding:

(f) software offered by the trader under a free and open-source licence, where the consumer does not pay a price and the personal data provided by the consumer are exclusively processed by the trader for the purpose of improving the security, compatibility or interoperability of that specific software;

Since this is already law in place and transposed, it would be a good starting point for consistency.

---

Publicpolicy mailing list -- publicpolicy@lists.wikimedia.org To unsubscribe send an email to publicpolicy-leave@lists.wikimedia.org

Hi Henning,

Just to put you at ease: I am not making such a proposal currently, but I want to discuss options at Wikimania and other community events.

I must however correct you on "Educational content is per se suitable for all ages". This is simply not true. Most educational is written with very specific age groups in mind. A three year old will learn from a specific category of books (most of them clearly marked with an age recommendation), while certain other topics (including sexual education, drugs, violence) are introduced in educational materials targeted at older kids. I know of no educational system in the world that doesn't apply some sort of age appropriate structure.

Dimi

Le jeu. 1 juin 2023 à 12:23, Henning Schlottmann h.schlottmann@gmx.net a écrit :

Hi,

Am 01.06.2023 10:17 schrieb Dimi Dimitrov dimi@wikimedia.be:

Another general solution I personally like is to move age-verification to the device or browser. [...] In this scenario the community would need to provide metadata/categorisation for sensitive content. Not sure it is feasible, but this is a universal approach that doesn't require the hovering up of user data.

I don't want to consider this. Educational content is per se suitable for all ages and should be exempt from any such demand. There is no age for which educational content is "dangerous".

And Dimi, please do not support any such proposal towards law-makers, because the communities will not follow you. I promise you, and everyone on this list, that all mayor projects will not categorize content for "appropriate age". I for one will prefer to shut down Wikipedia over censoring access to our work.

Henning _______________________________________________ Publicpolicy mailing list -- publicpolicy@lists.wikimedia.org To unsubscribe send an email to publicpolicy-leave@lists.wikimedia.org

Hi again,

I understand that the discussions around controversial content, especially on Commons, have never been easy and we have never managed to get to a consensus.Don't get me wrong, I would also prefer to not change anything. I am not advocating for content-gating solutions with lawmakers. But I want to have this very difficult discussion, not avoid it.

The world is changing and age-gating will be a huge legislative topic in the years to come. I guarantee you that. And if we want to continue hosting things like Category:Anal sex https://commons.wikimedia.org/wiki/Category:Anal_sex, Category:Pornographic_videos https://commons.wikimedia.org/wiki/Category:Pornographic_videos or even File:Close_view._Dead_Gaza_girl_day_14_of_Gaza_War_(2008–2009).jpg https://commons.wikimedia.org/wiki/File:Close_view._Dead_Gaza_girl_day_14_of_Gaza_War_(2008%E2%80%932009).jpg, it will become very hard at some point to convince lawmakers that educational repositories like Commons should not be obliged to age-gate.

My fear is that if we don't figure out an elegant way to convince people that graphic content on Commons is not an issue or don't give parents at least some control, we might get hit with harsh obligations that we don't like.

This is why I am trying to escalate a debate on this. Here, at Wikimania and elsewhere. We need to figure out some answers, as a movement, so we can advocate for them.

Cheers, Dimi

Le jeu. 1 juin 2023 à 12:59, Henning Schlottmann h.schlottmann@gmx.net a écrit :

Kids of reading age have free access to all content of the world and decide on their own what is of interest for them. This is true now for at least one full generation. And the result? We have the least violence between kids and the most peaceful generation since humanity exists.

Specific educational content, aimed at certain age brackets does not invalidates my claim, because this content is produced to raise interest in topics. I speak of intrinsic motivation to learning and all of us remember how we went far beyond "age appropriate" content to learn, when we were motivated. I want our content to be available to everyone, who is actively searching for and reading it.

As soon as I could read, my parents put the family 20-volume Brockhaus encyclopedia on the lower shelves in the living room. And I was fascinated by the images of the human body that developed from page to page from the outside of the nude body over a skinned body with the muscles showing, the inner organs and arterial system to the bare skeleton. When I had read almost all the children and most of the youth books in our small branch library I ventured into the non-fiction section and read voraciously through whatever I wanted that day. The librarians knew me by then of course, and occasionally had to override the computer for me to borrow books that were marked for a higher age bracket (the Munich municipal library system has long since abandoned marking books for certain ages but back then they did).

Please do not even discuss this with law-makers, the communities will not follow you.

Henning

Am 01.06.2023 12:33 schrieb Dimi Dimitrov dimi@wikimedia.be:

Hi Henning,

Just to put you at ease: I am not making such a proposal currently, but I want to discuss options at Wikimania and other community events.

I must however correct you on "Educational content is per se suitable for all ages". This is simply not true. Most educational is written with very specific age groups in mind. A three year old will learn from a specific category of books (most of them clearly marked with an age recommendation), while certain other topics (including sexual education, drugs, violence) are introduced in educational materials targeted at older kids. I know of no educational system in the world that doesn't apply some sort of age appropriate structure.

Dimi

Le jeu. 1 juin 2023 à 12:23, Henning Schlottmann h.schlottmann@gmx.net a écrit :

Hi,

Am 01.06.2023 10:17 schrieb Dimi Dimitrov dimi@wikimedia.be:

Another general solution I personally like is to move age-verification to the device or browser. [...] In this scenario the community would need to provide metadata/categorisation for sensitive content. Not sure it is feasible, but this is a universal approach that doesn't require the hovering up of user data.

I don't want to consider this. Educational content is per se suitable for all ages and should be exempt from any such demand. There is no age for which educational content is "dangerous".

And Dimi, please do not support any such proposal towards law-makers, because the communities will not follow you. I promise you, and everyone on this list, that all mayor projects will not categorize content for "appropriate age". I for one will prefer to shut down Wikipedia over censoring access to our work.

Henning _______________________________________________ Publicpolicy mailing list -- publicpolicy@lists.wikimedia.org To unsubscribe send an email to publicpolicy-leave@lists.wikimedia.org

-- Dimitar Dimitrov Policy Director Wikimedia Europe

mobile: +32497720374 Rue Belliard 12 Belliardstraat, Brussels https://wikimedia.brussels

Wikimedia Europe ivzw

---

Publicpolicy mailing list -- publicpolicy@lists.wikimedia.org To unsubscribe send an email to publicpolicy-leave@lists.wikimedia.org

Am 01.06.23 um 17:11 schrieb Henning Schlottmann:

Let's fight against it, with tooth and nails. Not cave in.

Sidenote: I'd find it interesting how the contributions of civic society organizations has shifted over the years as far as policy is concerned. Latest “compromises” e.g. with electronic ID on the net – of course always under the umbrella of constructiveness – find me a bit uneasy.

regards, -stk

On Thu, 1 Jun 2023, Spencer Graves wrote:

 How could we organize a maximalist campaign like Henning suggests 

with a reasonable chance of winning? And how can we combine that with a contingency plan that has a high likelihood of allowing us to disengage with minimal losses if it becomes clear that most likely alternatives involve much larger losses?

Here is what I propose:

We rephrase the matter at hand demonstrating some important aspects to the issue, for example (the list is not exhaustive):

- that children are autonomous subjects, not just objects to the parental control

- more often than not it is children that are in control of the Internet access at home, for example helping parents to set up their smartphones

- very important but hard to raise is the issue of discoverability: it is not true that we only deal with websites and "content", all of it looks like a shopping mall where people walk by, but there are many ways to carry out conversations ("transmit data") in the today's networked world. To make it very short, we have to distinguish between the pushed and pulled information. More often than not this will require limiting of "push" channels such us spam and advertisements. (Simplifying here a lot).

- jumping from something very general, to something very specific, I wonder how Among Us fared after introduction of pre-defined simplified "quick chat" in the app for kids under 13 in February 2021. From my personal observation (would love to find real data for that), this killed the app in that demographic immediately. I use this example to show that maybe we can have potential support not only among the usual suspects, like, for example, independent and SME game developers.

What we could do is to organize a series of events where kids/teens (invited to Brussels from all over the Europe, for example as a result of an on-line competition organized by us) actually demonstrate how their online life looks like, where are the dangers and what are the precautions they take (or not). Also, how they avoid all those filtering and content control measures. I find it fascinating how even young kids take measures to protect their identity on-line, for example.

(Some of you might have heard from me that I have attended a similar workshop organized by the European Commission many many years ago, bringing teens from a Polish high school to Brussels. It was amazing to see how the presence of well prepared young people literally smashed the efforts of the firewall lobby ready to offer products for parental control, everywhere).

I think we do not have to look far to find such people; we have probably the best ones in our community already.

Marcin