On 6 November 2026 the European Commission told TikTok that its “For You” feed preliminarily breaches the Digital Services Act (DSA). The reason: the way the feed recommends videos can pull users into harmful “rabbit holes” of content about sadness, isolation and body image. The finding is preliminary and carries no fine yet, but it is the third set of preliminary findings the Commission has sent TikTok this year, after addictive design in February and minors' account safety in July.
The Commission says the problem is not illegal content. It is repeated, personalised exposure to legal but risky material, and TikTok's alleged failure to assess and reduce that risk, especially for minors and vulnerable adults. Henna Virkkunen, the Commission's Executive Vice-President for tech sovereignty, said the system can lead users “down rabbit holes they cannot easily escape from” and that TikTok “needs to step up its efforts to protect users' well-being.” This article is for general information only and is not legal advice.
What the Commission alleges
The findings centre on both risk assessment and mitigation. In the Commission's view, TikTok did not properly assess how the For You feed can harm users, and the measures it has in place do not work well enough.
- The feed loop. The more a user watches or likes one type of content, the more similar content appears, including videos about mental health struggles and body image.
- The harm. According to the Commission, repeated exposure can damage mental health, especially for vulnerable adults and minors.
- Weak safeguards. Reports say techniques meant to diversify content do not sufficiently reduce how often similar videos are recommended, a feed refresh is only temporary, and the “not interested” button does little.
Changes the Commission wants TikTok to consider, as reported: relying less on implicit signals such as viewing time and more on explicit user preferences, making controls easier to find and use, making “not interested” stick, and testing the algorithm regularly. TikTok has rejected the findings. In a statement reported by AP it called them “fundamentally inaccurate” and said it gives users “direct and intuitive control” over their content experience.
Background: the DSA and TikTok's case file
The DSA gives the Commission direct supervision of “very large online platforms”, those with more than 45 million monthly users in the EU. TikTok is one of them. These platforms must assess the systemic risks their services create, including risks to mental health and minors, and put effective mitigation measures in place (Articles 34 and 35). The investigation these findings belong to was opened in February 2024.
- 6 November 2026: For You feed. The recommender creates harmful “rabbit holes”; risk assessment and mitigation fall short.
- 23 to 24 July 2026: minors' account safety. Minors can set accounts to public, and older minors' content can be recommended to anyone through the For You feed. Sources give 23 July and 24 July.
- February 2026: addictive design. Infinite scroll, autoplay, push notifications and personalised recommendations were not adequately assessed, and screen-time tools were too easy to dismiss.
What happens next
No fine has been imposed. A preliminary finding is the Commission's provisional view and does not prejudge the final outcome. TikTok can now examine the Commission's file and reply in writing, and the European Board for Digital Services, the body representing national Digital Services Coordinators, will be consulted.
If the Commission confirms the breach, it can adopt a non-compliance decision with a fine proportionate to the nature, gravity, recurrence and duration of the infringement, capped at 6 per cent of the provider's total annual worldwide turnover. It can also require TikTok to change how the feed works. We could not find a final decision in this case, so the timeline is open.
Why it matters for European businesses
The case shows that the Commission now treats recommender design as a compliance issue, not only a content-moderation one. Enforcement is moving from what users post to how a platform's own algorithm shapes what they see. Three takeaways follow. They are our analysis, not the Commission's.
- Algorithms are in scope. Any business that builds feeds, recommendations or engagement features should expect regulators to ask how those systems were risk-assessed and whether users can meaningfully control them. A personalised feed is also profiling under the GDPR, so the two regimes overlap.
- Evidence of effectiveness matters. The criticism is that TikTok's safeguards exist but do not work in practice. Documenting that a control is both available and effective is becoming the standard.
- Minors raise the bar. All three 2026 findings stress minors and vulnerable users, which fits the wider push on child protection online. See our guide on AI chatbots and children.
Smaller companies are not subject to the very-large-platform rules, but online platforms still carry lighter DSA duties, such as explaining the main parameters of their recommender systems (Article 27), and they sit alongside the GDPR and the EU AI Act in the same enforcement climate.
What to do now
- List every feature that ranks, recommends or auto-plays content to users, and note what personal data feeds it.
- Write down how you assessed the risks of each, especially for minors, and keep that assessment current.
- Give users controls that visibly work: an explicit preference should outweigh passive signals, and “not interested” should stick.
- Test and log whether those controls actually change what people see.
- Check whether your recommender relies on profiling, and update your privacy notice and DPIA if so. See our post on algorithmic management for how regulators view automated ranking.
The bottom line
The TikTok finding is preliminary, but the direction is clear: regulators will ask not only what your platform shows, but why, and whether people can steer it. If you run personalised feeds or recommendations, now is the time to document and test your safeguards. Book a free consultation with GDPRGard → and we will help you map the data and risks behind your recommender.
Sources
- Lovguiden: EU says TikTok's For You feed breaches the DSA over harmful rabbit holes (6 Nov 2026, repost of the Commission announcement)
- Associated Press via NY1: EU says rabbit hole video feeds on TikTok violate its digital rules (6 Nov 2026)
- The Next Web: TikTok faces EU fine over the For You feed rabbit hole effect (6 Nov 2026)
- MaltaToday: EU Commission finds TikTok feed breaches rules over harmful rabbit holes
- Cybernews: EU finds TikTok's recommendation algorithm in breach of DSA
- Tech Insider: TikTok rabbit hole feed and the 6% fine cap
- IDPC Malta: Commission preliminarily finds TikTok in breach of the DSA over safe accounts for minors (July 2026)
- Digital Services Act, Regulation (EU) 2022/2065, EUR-Lex
Pages used for this article as of 11 November 2026. The Commission's own press release was not read directly, so quotes and details come via secondary outlets and a repost of the announcement. TikTok's response comes from the AP report.
Also read:
- Built for Everyone, Used by Kids: What Europe Now Expects From AI Chatbots
- When the Algorithm Is the Boss: AI at Work After the €825 Million Uber Fine
- Your AI Agent Can Act on Its Own. Privacy Law Still Expects You to Be in Charge