The European Commission has put Meta on notice: change how Facebook and Instagram keep people watching and scrolling, or risk penalties under the Digital Services Act.
On Thursday, the Commission said its preliminary investigation found that Meta failed to properly assess how design choices on the two platforms may affect users’ physical and mental health, including minors and vulnerable adults. The features named by regulators are not exotic: autoplay, endless feeds and recommendation systems tuned around engagement.
The Commission said those mechanics encourage users to keep scrolling and can push them into automatic, compulsive use. That is the policy problem for Meta. Under the DSA, very large online platforms must assess systemic risks and take steps to reduce them. Brussels is saying Meta’s current answer does not pass.
If the Commission reaches a final non-compliance decision, Meta could be fined up to 6 percent of its global annual turnover. The company will be able to contest the preliminary findings before a final decision, expected in the coming months.
What Brussels wants changed
The Commission said Meta should consider disabling autoplay and infinite scroll by default, adding more effective screen-time breaks and changing its recommender systems so they are less centered on maximizing engagement.
Henna Virkkunen, the EU’s technology chief, told Reuters that the Commission’s starting point is that the design is too addictive and must be changed. She said Meta can either change the design or face a non-compliance decision.
The Commission also rejected Meta’s argument that its existing protections are enough. Regulators said teen-focused time management tools enabled by default did not adequately address the risks created by the platforms’ design. They also said parental controls depend on parents or guardians having enough technical skill, time and effort to use them effectively. That is a fairly polite way of saying a settings maze is not a safety system.
Meta disagrees. Ben Walters, a company spokesperson, told Reuters that the preliminary findings do not properly account for the steps Meta says it has taken to protect teenagers. Walters pointed to Teen Accounts, which he said automatically add protections, let parents block Instagram access at night and allow daily screen time limits as low as 15 minutes.
Pressure is building outside Europe too
The EU action lands while Meta faces scrutiny in the United States over similar allegations. Reuters reported that Meta recently failed to dismiss claims brought by 29 states alleging that Facebook and Instagram addict children. That trial is scheduled to begin in August, and Reuters reported that states may seek up to $1.4 trillion in penalties if they prevail.
California Attorney General Rob Bonta told Reuters that Meta prioritized profit over child safety and contributed to a mental health crisis affecting American children. Meta disputes the broader claims that its platforms are unsafe for teens.
The Commission also said expert findings expected Monday could help inform debate over a Europe-wide social media ban for teenagers. Several governments have already passed or are considering restrictions on minors’ access to social platforms.
Meta is dealing with the fight while pouring money into artificial intelligence. The Information reported that Mark Zuckerberg has allocated between $125 billion and $145 billion for AI data center capital expenditures this year, alongside higher operating costs for cloud services and AI talent. Business Insider reported that Meta told employees an upcoming model, codenamed Watermelon, had caught up with OpenAI’s GPT-5.5, citing people familiar with an internal meeting agenda.
Another Meta AI release has drawn privacy criticism. NBC News reported that Meta’s Muse model uses public Instagram photos and videos and can be used to generate deepfakes that Meta’s detection tools do not always catch. Meta said most users were opted in by default, except users with private profiles and under-18 users whose sharing and reuse settings were off by default. The company told Reuters that users can opt out in a few clicks.
For EU regulators, the immediate issue is narrower than Meta’s AI spending or product roadmap. They are asking whether Facebook and Instagram’s core mechanics are designed to keep users engaged in ways Meta has not justified or mitigated under European law. Meta now has to answer that in Brussels, not in a product blog.
This story draws on original reporting from Ars Technica.
