- “And, um, and these each of these, um, are a mix of rights and obligations. Uh, and they are different. And so it's just important that in as we speak of the different concepts that we delineate quite clearly, um, whose rights and obligations we are speaking about. And just to reassure everybody, we are keenly aware that the duties of the European Commission as enforcer are constrained by the Charter of Fundamental Rights, and notably, the rights therein of, uh, freedom of expression. But I also want to stress, um, that, uh, in the European context, um, it is not only the speakers that have rights, but also the audience. And uh, because the charter says that it is a freedom to impart and receive information and that that, uh, is different, I think, from other jurisdictions and is something important to take into account as we implement the rules. So with that, I yield my time. Thank you very much.”
Digital platforms liability for harmful and illegal content · Disinformation & online freedoms
- “Yes, thank you Madam Chair and thanks to all the MEPs for these tough questions. You know, these are not easy, but I'll give it a shot. Maybe first to the question on do social media bans unduly restrict children's fundamental rights. I think so, just to be perfectly honest. I think that this is a question we have ourselves and we will put into the expert processes. I think the extremes are very easy to understand. It depends on exactly how the ban is constructed. Now, if you prohibit all platforms for anyone below eighteen, I think that is probably not consistent with children's rights to access some informational rights that they have, the rights to play and to learn, and so on. Of course, nobody's proposing this, I just wanted to illustrate that if the restriction is excessive, then it will negatively affect children's rights. At the same time, no restrictions at all, in our view, and this is also the content of the guidelines, actually do more harm than good. This is why we recommend that in a variety of different cases, we need at least age assurance. In some cases, like pornographic platforms or gambling or alcohol sales, we need also stricter methods as well. It's precisely said, it's a little bit the question, the answer also to, I mean, if you want to talk, I mean, there needs to be proportionality in the system. The areas of proportionality that we have put into the guidelines are, first of all, what is the actual risk of the system. I'll answer the question on why we think that social media are not the same level as pornographic websites in a minute. So there needs to be proportionality at the level of what is the inherent risk in the platform itself. Secondly, there's also another proportionality which is really important: do you always need to restrict the entire service or is it about parts of the service that are risky for children? So, if you have, for example, you can imagine a platform that has certain parts which are very high risk, then it is in the guidelines that we only restrict those parts that really have the risk to children and others could be left more unrestricted. The answer is a little bit a question of proportionality and it's hard to give a simple answer to what the effects of a ban are. Maybe also just to give the example of Australia, one of the controversial parts that we have been witnessing is whether platforms such as YouTube should be in scope or not. In the end, I think there was control and discussions about this precisely because this is also a platform that's used by kids to do their homework if they haven't understood the math lesson in the afternoon. We don't have a position on this right now, but I just wanted to say that this is part of the reason, this is part of the questions that we had and that we are aware of as well. On the second question, which was how do we ensure EU-wide uniform application of these rules, I think I'm very encouraged by the work together with national authorities that we are doing. First of all, I think there is a degree of uniform application in the DSA that is very strong by the Commission being kind of a common enforcer for very large online platforms, setting de facto uniform rules. We don't have that in many other areas. So if one set of rules and one single enforcer for these rules, I think that also helps tremendously. Then what we have seen and heard in our interactions with national authorities that are enforcing, there's a huge amount of willingness to apply the guidelines also for smaller platforms in a consistent and coherent manner. This was very much the focus of our discussion yesterday. To the extent that, of course, also the DSA is a federated enforcement system across the European Union, like every piece of EU law actually, it has more harmonizing elements than other pieces of laws because of the Commission's role in enforcing against very large online platforms. I actually think that I'm really encouraged by the common approach that we are taking here. We are also pursuing joint investigations and enforcement approaches against the pornographic platforms, many of which are smaller than the forty-five million level threshold. On the questions that were asked about the age methods and the technology of these age verification methods, I think we still intend to come forward with a technical document on the precise assessments on which technologies satisfy the different criteria that are set out in the guidelines. We have these criteria on robustness, nondiscrimination, and so on. Our preliminary thinking is that the EU digital wallet solution is really one of the maybe only solutions that meets all of the criteria, including the privacy-preserving one. What I just want to say really super clearly is having platforms and social media platforms scan and upload government ID cards is not consistent with the guidelines. This is not what we, you know, this is something we've seen in other jurisdictions. Some platforms have proposed it, but this is not consistent with what the guidelines say. So I would say the EU digital wallet solution is a solution that meets all the criteria set out in the guidelines, including the privacy ones. But of course, it's also a complex architecture. It requires somebody to give credentials that are age credentials. It requires an app or another infrastructure that can pass on these credentials to the platforms. It needs many different parties to work. This is actually the objective of our pilot phase: to make sure that all of these different parts work together. I think we are making progress. We have seen five or six member states that are piloting this and integrating this into the national digital ID infrastructure. We will be reporting on this pilot here also with pleasure back to this committee as soon as we have tangible results to show. On the waterbed effect, I think yes, there are many waterbed effects here. There's the waterbed effect of children moving to smaller platforms that are less regulated and jurisdictional escape. There is a waterbed effect about less content moderation if children are on those platforms. But I think the rules of the DSA will stand. Also on other harms, for example, the rules on addictive design are not just limited to minors by the way in the DSA. While I see your concern and we are probably vigilant for doing this, including the risk of circumvention, I think we will have to continuously monitor this and then take corrective actions once we observe it. But I think it cannot be a reason not to do this step right now because something bad might happen. By that logic, we wouldn't be doing very much. I think you very much.”
Privacy & detection of online child abuse · Digital platforms liability for harmful and illegal content · Safety features & content control for child protection online
- “Thank you. Yes, thank you also for this question. The only legitimate reason to request access is to research EU risks. You cannot, nobody has any rights to research risks outside of the European Union. However, the law does not specify who in the world can access this research. We have included in the delegated act some indications. One of the indications that is implicit is that if personal data is involved, then it cannot be transferred out of the Union unless there are safeguards in place in law and under our international transfer rules. This is one point. Second, in practice, what will happen, and this is also our experience, is that the consortia is not usually one researcher, it's a group of researchers. They will de facto be anchored in the European Union. This is just practical because you need to interact with a national regulator and comply with a whole set of requirements. For somebody outside of the EU, even if it's theoretically possible, the barrier of demonstrating that they have legitimate access is very high. It's not impossible but it has to be for risks inside the Union. We believe the threshold is very high.”
International data transfers · Privacy & law enforcement
- “Thank you, thank you very much Anna and always a pleasure to be in the committee. Also, thank you to Laura Porter and her team as well as the interventions. I'll be very brief and hand over to my colleague Mirto just in a minute. As you know, protecting minors online is really a key priority of the Commission. It's something that we have now even set up a dedicated team for inside the DSA enforcement team. It's essential that minors' privacy, safety, and security are well protected in the digital age. We therefore warmly welcome the work on the report. It shows that we really share many of the priorities around protecting, promoting, and respecting children in their rights and their rights online. We also welcome a recognition in many amendments of the work that is being done on this already, notably with the Better Internet for Kids strategy and the Safer Internet Centers, the Audiovisual Media Services Directive and its upcoming evaluation and possible review, but also the Digital Services Act with its enforcement and guidelines, the age verification blueprint that has also been mentioned by some MEPs, and the Cyberbullying Action Plan, just to name a few. Let me just respond to three elements in the draft report and the amendments. One on the calls for swift and better DSA enforcement. This is something I think about day and night more or less, and it's also the reason why we have restructured our teams recently on the first of July to actually be able to better distribute the workload around enforcement actions and to be more effective, faster, more efficient in our investigatory actions. We have taken action recently against four pornographic platforms: Pornhub, Stripchat, XNXX, and XVideos for suspicion that in violation of the Digital Services Act, minors have unfettered access. We're also running other cases focused on protection of minors, including algorithmic systems, addictive design, rabbit holes of harmful content, as well as the efficacy of age checks to protect minors' rights and mental health online. We will continue to pursue these cases with the utmost urgency. I have already mentioned previous enforcement cases to this committee; I won't go into too much detail. Last year, we took action against TikTok Lite and extracted commitments which actually also addressed directly the issue of online addictive design. The guidelines that have been mentioned have been only recently released, outlining a very ambitious set of principles that we will use in the enforcement of the Digital Services Act. These are not just empty letter guidelines and these are not just nonbinding words, but these will guide and direct the expectations around the enforcement of the Digital Services Act. The blueprint for age verification is separate and different from what many other jurisdictions have been doing in the space of age verification because, precisely in response to the committee's concerns, it offers the possibility of completely anonymous verification of a user's age, unlike the practice of scanning ID cards by the platforms themselves or similar features that would be very intrusive. I can report here that the first results of the pilot engagement with the providers as well as Member States' authorities have been encouraging for us, and we have good results with the technical teams working together across many of the platforms. Let me also mention briefly that on the interaction with other legislation, in November the Commission will publish its report on the articulation of the DSA and other laws. This will include some of the legislation that's mentioned in your report. I'll be making just two small points now. One on the calls for further limits or restrictions to children's access. You have heard President von der Leyen from the lines' emphasis in her State of the Union speech here to the European Parliament to empower parents, protect children better, and the proposed exploration via an expert group of possible further restrictions for social media access. To this end, we are setting up an expert panel by the end of this year to advise on the best approach for Europe. Indeed, it's not just because we are front runners on many aspects that we count to more have to listen to everyone, of course, from experts to parents to Member States and, of course, to children themselves as well. We have very close cooperation agreements with Australia and are following very closely everything that is going on. Of course, I wish to recall that we also are engaged with our colleagues under the leadership of Commissioner Várhelyi and jointly, together with EVP Verkuenen, on an EU-wide inquiry into the impact of social media on mental health and well-being. On the calls for further legislation, you know my position and it has been always repeated here that we, of course, are ready to work with you to identify any possible gaps. We would just warn and caution against creating additional overlapping or conflicting rules as well. I'll hand over to my colleague Milto at this point.”
Privacy & detection of online child abuse · Digital platforms liability for harmful and illegal content · Safety features & content control for child protection online
- “Thanks. Um, what I can say is that, um, we are seeing, um, in the work that we are doing indeed, the transformation that has been described by several speakers, uh, of the increasing importance of artificial intelligence for, um, the use of recommender systems and content moderation systems by um, the regulated platforms, notably the rise of transformer technology, uh, and a shift away from previous uh, recommender systems, which were more, uh, technology More probabilistically based. Um, this, uh, changes, of course, the risk landscape. Uh, and, um, and it is, uh, work to understand exactly how, uh, the risks associated with, uh, recommender systems are being transformed in that context. I'd just like to highlight that we have ordered, uh, notably on, uh, the platform X access to, uh, the, uh, history of documentation on the changes to the recommender systems back in, in January 2025. And, uh, along with access to certain, um, documentation, documentation. So I can also confirm that access complied with these orders, uh, and has submitted, uh, a very large volume of documentation to the European Commission. My third point, um, is just, uh, that, uh, um, we need to be and I think this was said by many speakers quite careful in distinguishing between, um, the, um, the rights of the speakers in the online environment, the rights of the audience in the online environments, the obligations on online platforms, and the duties of the different regulatory bodies that are there.”
Digital platforms liability for harmful and illegal content · Recommender systems
- “Yes, thank you very much. Good questions all around and pleasure to try and answer as much as I can. First of all, the question: how does the mechanics work? Are there national authorities everywhere and do we have enough vetted researchers who can actually make use of these new rights? We have now authorities in every member state that have been named but not in every member state have they been fully legally empowered. I actually have to confess that I don't know the absolute latest numbers but we have opened some infringements against a number of countries where the final legal powers have not been given. We have taken those countries also to the Court of Justice very quickly. By the way, I have to say that these infringement procedures we did not wait and they are already at the last stage, these infringement procedures of the referral to the court. The Cyprus authority was the last one that has just been empowered but I think we are talking about a very small number of authorities that are not yet fully active and activated and we're pursuing them in court. I think the system is almost there and I think it's faster than if you compare to other laws like GDPR or something like that. Are there researchers everywhere? I think there are some specialized researchers everywhere that understand what we are doing but the number is small and needs to grow. We are really working very hard. Today actually my colleague is in Berlin for a big workshop explaining how these routes work with researchers. We are working with our own research funding bodies like I mentioned the European Research Council to create awareness and provide practical help as well. We have done this piloting work with some researchers and they talk to each other of course. I don't think that from day one we will see applications everywhere but I do think that we will see a good number of applications as well. I think the system is as well prepared as it can be at this particular stage but we will try and advance as well. You had a question also on the safeguards by the platforms. Do they have to give all of their data? No, the rules are set out very clearly and the platform can also refuse but there is a process. They cannot refuse blatantly inventing some reason. There are clear criteria on when the platform can refuse and it will be the national authority that decides. It's not just the platform; it cannot just say no like this. There has to be a reason to request and actually when they say no they have to propose an alternative. There is a process and then it will be the authority to decide to facilitate this and to avoid that we have too many rejected applications because of this ping-pong process. The delegated act also sets out some more detail on which kind of datasets we expect to be easily available by all platforms. There is some guidance contained in the delegated act that says if you get a request for this kind of categories of data, we really assume that you have this data and you should get ready in making these datasets available as well. That's a little bit the mechanics. Trade secrets are protected. Platforms can make a reasoned rejection. It will be up to the national authority to decide but it cannot be an excuse to frustrate the process. This is a little bit the way we designed it. If I take it in terms of the general and the specific point, what change is actually happening and are these enforcement actions really changing the environment? I have two answers. One is that one of the features of the Digital Services Act and regulation in this space is that what you see is what's left. The problems that get corrected disappear from view because they have been addressed. One part of our work where we can do better is to explain some of the positive changes that have come about. Many platforms have done many good things in response to compliance with the Digital Services Act. Public attention naturally focuses on those things that still remain problematic. It's less common to speak about the problems that have been solved but in many areas there has been progress. I can give concrete examples. One of the big asks of the Parliament was to stop targeting minors and even targeting anyone with sensitive data categories on ads. This just happened. In the one or two cases where we had doubts, platforms reacted very quickly. It's easy to forget that this was really controversial at the time and that this has just been eliminated. To the degree of our ability to see this and monitor this, we had one civil society complaint on which we acted and we were able to get change very quickly. When we were negotiating the Digital Services Act, targeting with sensitive data categories for advertising was a really big deal but it has just been removed from view. Of course, now we focus on new harms that have arisen but I want to say that success of the DSA should also be measured against many of the things that have actually come about. We can do better in communicating some of this together with platforms. On the speed of investigations, I think we will see an increase in pace of enforcement decisions in the next phase. Partly we had to build a regulator, create structures, and be sure we are not making procedural mistakes. In an area where procedural mistakes are costly, they can cost a case. I'm confident as a responsible official for many of these decisions that are in the pipeline, you will see an increased pace of delivery in the next weeks and months. On the distinction between publicly accessible datasets, this was a question asked by two MEPs, and internal datasets. There are two provisions: platforms have an obligation to give you data that you and I can see as well. If you go onto the social media website, you can see somebody's profile picture, data, and history, then a researcher should be able to see it as well and acquire data even at large scale. This is what we mean by publicly accessible data, a rule in Article 40, paragraph 12. This is also where the CrowdTangle question comes in. We have six enforcement cases open and have taken commitments from AliExpress which are very far-reaching. Other cases are progressing well. Basically, if you and I can see some content, then this content must also be accessible to an independent researcher. Of course, these researchers don't look at individual things but want to look at large-scale effects. Some of our cases focus on whether this is done or not. The other provision is on internal data access to internal data of the platforms themselves. That's stuff you and I cannot normally see but for example engagement metrics, how many people engage with particular content or view metrics. They are not always public but are really important in understanding whether some conduct is risky or not. That's the scope of the delegated access act for data access right now. I think this is a really important piece. We have prioritized data access in our enforcement work because it was so important to everybody and the research community. It's quite revolutionary. There's no other area, neither nuclear energy nor medicine nor pharmaceuticals, where researchers have access to privately held datasets. It's quite innovative and also a reason why it takes a bit of time to develop but I'm confident we are getting there. Thank you very much.”
Interoperability requirements for digital platforms
- “Thank you very much Anna. I'm looking for my speaking points but I think you will have had enough of me at the end of the day. This is actually one piece of work on the DSA which I'm also very proud of. Again, I want to just call out the teams that have been working on this. This is the piece on the delegated regulation of vetted researchers' access to data. This is also something that I know this Parliament has been fighting for in negotiation of the Digital Services Act. I'm glad to give you an update on where we are with this particular provision. If we just look back at the previous two agenda items and the kind of questions we've been asking, also is the content consumed by children impacting their well-being? This is a question we've just been discussing. I mean, why do they see what they see on their personalized feed? If you take the personalized feed which most kids use as default, on the 'For You' feed whether it's on TikTok or Instagram or any of the other platforms that are commonly used. How can we assess if the measures that platforms themselves are implementing, regardless of the guidelines, are actually reducing the risk for children when they are on social media? Of course, these are questions that we as regulators are looking at, but what you always need is data that is actually in the hands of the platforms themselves. It's impossible to answer many of these questions without looking at platforms' own internal data. This is the reason why the DSA has this very revolutionary provision, in my view, that allows independent researchers to carry out research on systemic risks by getting access to platform internal data. Our experience has shown that very often platforms' own analysis of risks that materialize on their systems have many blind spots. In a system where platforms are marking their own homework, this becomes risky because such blind spots may lead to real-world harms. This is why we have this provision on solid, reliable third-party research to get information about what works and what doesn't work in the online space. July was a busy month for us. We published these guidelines but we also took enforcement actions against Timu. I know this is not a subject today in terms of e-commerce marketplace, but I know this committee is often very interested in what we do. The third big deliverable in July was the adoption of the much-awaited delegated expert data access to data by vetted researchers. Now we are in the transition period until October when these rules become binding. There are clear rules on how the access modalities are going to be implemented. A lot of work has to be done by national authorities who will play a key role in the success of these mechanisms because they are ultimately the guardians of the procedure and will protect the rights of the companies but also the interests of the research community. Only if a national regulator is convinced that the researcher can actually perform the research properly will a request be granted. This is one of the important safeguards to protect also the data of the companies involved. We have taken a lot of time in balancing the rules that protect the researchers and their interests with the rights of the companies as well. We have been extensively consulted. This is something completely new. We've run pilot projects with platforms, authorities, and some of Europe's leading researchers on prototype rules and mechanisms on how it would actually work. This is a kind of road testing that we did, one of the reasons why it took a little bit longer than initially planned on getting these rules out. Researchers are now gearing up, started preparing their applications. It's also clear that such applications require a little bit of a mind shift in the way research projects are devised because there are some legal criteria. We try to lower the barrier as much as possible for researchers' access but there is some due diligence that needs to be fulfilled. A couple of additional points: in getting ready for this vetted researcher, we have worked very closely with the European Research Council, which is the European Union's top research arm. We've worked carefully to make sure there is good awareness amongst European researchers on how these rules are working across a variety of different areas including mental health and many other areas of work. We look forward to good engagement with European scientists to make use of the new rights that have been granted. Next to this new delegated act for vetted researchers, we are continuing our enforcement actions also on access to publicly available data on Article 14. We have a number of cases open because transparency is the cornerstone on the basis of which other actions have to be built. Six services have been undergoing investigations for suspicions of noncompliance in this important provision. One thing to call out in July was that the Commission accepted a series of commitments proposed by AliExpress to address failures identified in the system that the company made available for maintaining effective data access for researchers. We have other transparency tools including the database for keeping track of content moderation decisions in real time. It has more than thirty-five billion moderation decisions right now since September 2023. Thanks to this tool, we know that there's an absolutely negligible tiny fraction of content moderation decisions that are linked to any kind of government requests taken by the platforms. That's also a really important piece of transparency that we are doing. Let me conclude: just like the medical research community has harnessed the power of medical data for the benefits of patients everywhere, the social media research community can now also harness the power of social media data to improve our understanding of our lives online and help to make safe services safer in the process. We really look forward to successful data access applications and we're working very closely with the national regulators to ensure that this becomes a reality as quickly as possible. Thank you very much.”
Disinformation & online freedoms · Digital platforms liability for harmful and illegal content · Safety features & content control for child protection online
- “Thank you very much Anna. I really welcome the opportunity and I know I come and speak here but I just want to also do justice to my team that has been working really hard on not only the guidelines but also all the work that has been going on in the background. I think in a way it's an area where we have actually moved incredibly fast with international standards as well, where similar guidelines have been developed in other jurisdictions but over much longer time scales than our own work. Before I take too much of our time, I just wanted to also do justice to all the colleagues, including some of which are here in the room but also others in Luxembourg, and the big teams that have been working really day and night also on the digital wallet solution, the Better Internet for Kids, the Safer Internet Centers, and the content of the guidelines as well. I won't present all the recommendations in the guidelines but I'll give you a high-level overview of some key considerations. First of all, they are a piece of the puzzle of a comprehensive EU-wide approach to protecting minors online. This includes the entire set of DSA provisions which don't only deal with Article 28 and the protection of minors, but as you know and as you have negotiated here in this committee during the DSA, there are other parts of the DSA that also aim at protection of minors. So this is one part but a very important part. Of course, next to it there's also the Audiovisual Media Services Directive and their protections that are in there, the age verification blueprint I already mentioned, the strategy, the Safer Internet Centers, the Cyberbullying Action Plan. I won't repeat them all as I've just mentioned many of them as well. We believe the guidelines are really the most ambitious, detailed, and comprehensive instrument anywhere in the world. They reflect our alignment and growing cooperation also with like-minded partners, notably the Australian eSafety Commissioner, the UK's Ofcom, but also as we have seen the US Federal Trade Commission that is active on this specific topic. Several other countries are advancing around the world including Brazil and Canada. On the content itself of these guidelines, they really aim to reflect the ambition of this Commission and the Executive Vice President Verkuenen who has made this a top priority when it comes to enforcing the Digital Services Act. They aim to assist online platforms in ensuring a high level of privacy, safety, and security for children. Particular, they adopt a risk-based approach and this was already mentioned by several members here in the previous discussion, recognizing that different platforms have different risks so one size cannot fit all. It may not be proportionate for all platforms to apply all the measures set out in the guidelines. Conversely, there is a rich set of obligations in the guidelines that apply to many at the same time. They are based on children's rights because this is an area where normative rights have already been formulated in international human rights law. Our approach is fully consistent with that. Platforms ensure that the measures that are put in place do not unduly restrict children's right, for example, to participation or freedom of expression, which was also mentioned here as well. Safety by design is really at the core. This is something you do before you're designing a service and not only as an afterthought. Privacy, safety, and security are built in and baked in when the platforms are designed, developed, and updated, not just as a corrective after the fact. The guidelines apply to all platforms that are accessible to minors, apologies, so not just the large ones but to all platforms. They provide recommendations on all relevant areas in our view. Age assurance was already widely discussed. It's linked to the EU-wide age verification blueprint. It gives guidance on when age verification is required. It sets out minimum criteria for all age assurance methods, including a number of criteria that intend to guide the market as private sector solutions advance. Recommended systems have already been mentioned as well: account settings, default account settings for minors, interface design, design practices, notably those design practices that aim to reduce risks of addiction from features such as endless scrolling, streaks, or others, as well as dark patterns or other manipulative, addictive, and harmful design practices. These are really all summarized in the section on recommender systems as well. Content moderation, reporting tools, user support tools have to work in practice and need to be age appropriate. That means they have to take into account the specific needs of different children. Commercial practices including specific features like in-app purchases, virtual currencies, and loot boxes, as well as platforms' internal governance and tools for guardians, are also part of the toolkit set out in the Article 28 guidelines. Maybe I can just say one thing before we continue on the particulars. The letter of the law says in Article 28 that platforms have to ensure a high level of privacy, safety, and security for children. What these guidelines basically do is describe concrete steps that can ensure high levels of privacy, security, and safety for children online. This is really the reason why they are very elaborate because depending on the specific circumstances, different tools and different combinations of tools must be ready. The adoption was the beginning. Now it has to be followed up with concrete next steps including swift action on the ongoing investigations but also effectively enforcing against other very large online platforms under our supervision, but also to provide close collaboration with national authorities in charge of enforcing these rules at national level. We've just had an extensive discussion with the Board of Digital Services Coordinators only yesterday on the implementation of the guidelines. There's a really strong and uniform commitment from all national authorities in enforcing these guidelines also against smaller platforms as well. I'm also pleased to report that a number of national regulators are envisaging or have taken enforcement actions in this regard, also at platforms that are regulated at national level. Now on the relation of the guidelines with the potential social media restrictions that are legislated at national level, I just want to say that the priority here is to make the online environment safer. The DSA as a law that regulates rules on platforms, so obligations on platforms are regulated exhaustively in the DSA, does not allow national law to impose additional or different obligations on platforms. The guidelines recommend that the use of age verification, if this is done in a national law, needs to be done without interfering with this maximum harmonization method of the DSA. At the same time, the DSA does not set a minimum age to access services. This is the space where national laws can regulate as long as they don't have rules that contradict the DSA. Examples here, this may sound very complicated but examples just to make it clearer for you, is that national laws already regulate a minimum age for, for example, alcohol sales or cigarettes or vapes. A recent example in the Netherlands, very prominent, or access to pornographic content or gambling. This is already regulated at national level. Many Member States have rules that below a certain age you're not allowed to access such content or such services. So if there is a desired national level for a minimum age, this can be incorporated in the architecture of the DSA. Now when we go further and I'm almost coming to the end of my intervention, we have focused now on the guidelines but the guidelines also recognize that a wider context exists and risks of circumvention as well as the potential risks that their rights might be limited or even their ability to develop digital literacy. Excessive restrictions have to be weighed also against the risks of cutting children off from some of the experiences online that might be necessary for a good development in our society. So the decisions that should be taken in this area have to be based on evidence in our view. This is also why we welcome very much all the work that has been done here and will be done to inform wider evidence-based debate. We have therefore also started work on establishing an expert group on these restrictions in addition to the expert group on the mental health risks for minors. In doing so, we will be drawing on the work of this committee as well but also work that has been done at national level in many countries. I end by saying that the guidelines are not a goal but a starting point. They will have to be followed up by concrete and effective and visible action. They are the result of a granular, inclusive, and comprehensive effort to listen to everyone's voice including that of children but also of national authorities, industry, parents, and academics. We will continue this open dialogue with everyone. We will come forward with additional guidance on age estimation methods which is work right now as well as methodologies for the risk review for these issues. Work is ongoing as well because while the guidelines set out high-level principles, further details can be provided on which methods in our view satisfy which of the criteria as they are on the market today. Of course, this is an evolving area as well. My final point is that we appreciate that the sector changes very quickly and so we have committed ourselves in the guidelines to reviewing progress already next year, which is exceptionally fast for the work that we are doing, but we recognize that this is a quickly changing environment as well. With this, I end Anna and hand back the floor to you. We welcome, of course, continued engagement with the Parliament on this crucial issue and we stand ready to assist the committee for any further work that may be necessary. Thank you very much.”
Privacy & detection of online child abuse · Digital platforms liability for harmful and illegal content · Safety features & content control for child protection online