Children deserve a safer internet, said European Commission President Ursula von der Leyen when presenting the EU Kids Act on 17 September. The proposal sets graduated thresholds for social media access, bans addictive design features like infinite scroll for under-18s, and requires platforms to verify age using a privacy-preserving “yes/no” token instead of a birth-certificate check.
Commission Executive Vice-President Henna Virkkunen presented it as a delay, not a ban. Children get eased into risky platforms at 13, 15, and 18 rather than being locked out entirely.
That framing worries digital rights groups. “The way the proposal is framed is that age is the problem,” says Simeon de Brouwer, policy advisor at European Digital Rights (EDRi), a Brussels-based network of digital rights NGOs.
He argues that by delaying rather than fixing the underlying design, “the Commission does not address the causes of the harm, the business model and the addictive design, the manipulative practices, which young people will be exposed to as soon as they reach the different thresholds.”
Age as a proxy for vulnerability
De Brouwer’s core objection is that the Act treats age as the only marker of vulnerability worth legislating for. “The sole factor of vulnerability that is taken into consideration is young age, not even age. But there are many other factors,” he says.
EDRi represents member organisations working across civil society, and according to de Brouwer, the organisation tries “to take into account marginalised groups whose voice we don’t hear that much in Brussels.”
That matters because age verification requires everyone, not just children, to prove they are old enough to pass through the gate. “If we really need to know who’s underage to protect them better, that means we know everybody’s age, or we need everybody’s age,” de Brouwer continues. “We are turning smartphones and identity documents into prerequisites for accessing online spaces and exercising fundamental rights.” He lists who gets caught in that net: “people who don’t have state identities, people who are stateless, people from migration backgrounds and not integrated yet, people who by choice don’t have a smartphone.”
Who could be marginalised?
The same restrictions risk harming some of the young people the Act is meant to protect, including those with disabilities who rely on the internet to socialise.
De Brouwer points to “young people who also are marginalised, so people who are handicapped and don’t have that many physical interactions, people who are in rural spaces and don’t have friends around, people whose parents are not very present or willing to support them and therefore need online spaces to find, to talk to, to find people like them.” He concludes that “the range of young people who will lose out is immense.”
Francesca Pisanu, senior policy officer at Eurochild, a network focused on children’s rights, broadly welcomes the Act’s ambition but agrees the accessibility question deserves scrutiny. She says: “It is known that social media platforms or the online world in general is a space where children, or people in general, can be included in society. They can find people with shared identities succeed. They can also see more diversity.”
She adds a caveat, “people that are discriminating [in] the offline world are also likely to be discriminated [against] in the online world,” so it’s important that “all these inclusive features” are “activated at the same time as the protective features,” so the two do “not contradict” each other.
Pisanu pointed to a Eurochild-linked study in the Netherlands that found that “children with disabilities or the children with mental health issues were the ones that were actually in favour of a ban,” a reminder that marginalised groups do not speak with one voice on this question.
Overlapping and overriding other Acts
For de Brouwer, the deeper problem is political timing. He argued the Kids Act risks overriding the Digital Fairness Act (DFA), a parallel initiative meant to strengthen consumer protection “everywhere, regardless of which service you use”, not just for children.
“The risk is that we lose out by focusing so much on children and young people. We lose the political momentum that we had to improve everyone’s protection through the DFA,” he said, warning it could now arrive “in a diluted fashion.”
He was equally clear that many of the Act’s protections should not be limited to minors. “Who doesn’t want to be protected from AI companions designed to be addictive and to entice you to keep coming back?” he asked, arguing that platforms should instead be “forced to be non-addictive” and that “we should have, for example, a right to services which are not personalised.”
Pisanu’s institutional position is less confrontational: Eurochild wants safety-by-design to be binding law, not just DSA guidelines, and welcomes that some protections, like recommender-system defaults, would extend even to unregistered users, since some of these “safety features would apply also to those that are not registered.” But both interviewees converge on one point: the current draft answers “how do we protect children” without fully answering “how do we protect anyone who is vulnerable but not underage.”
The European Parliament and the Council must take positions on the Kids Act, and the Commission must review the rules’ effect on fundamental rights by 31 August 2030, leaving the balance between protection and exclusion unsettled.