Preliminary findings: the EU says TikTok violated the DSA by not preventing its recommendation algorithm from trapping users in harmful content "rabbit holes"

The EU is directly challenging the engagement-based recommendation algorithms of major social media platforms, a model that could fundamentally change how content is surfaced.

The European Commission has issued preliminary findings accusing TikTok of violating the EU's Digital Services Act (DSA) by failing to prevent its recommendation algorithm from trapping users in harmful content "rabbit holes." Specifically, the algorithm allegedly pushes users into endless streams of content related to sadness, self-harm, and body image, even when users interact with such content. Officials stated this is the first time EU regulators have directly challenged a major social media platform's engagement-based recommendation system and the first enforcement action of its kind globally. While TikTok acknowledged some risks in its own assessments, investigators found the company did not adequately analyze them or implement sufficient safeguards. The findings are preliminary, and TikTok has the opportunity to contest the evidence. If violations are confirmed, TikTok could face a fine of up to 6% of its global annual revenue.

This enforcement action signals a significant regulatory push by the EU to address the design choices of social media platforms that prioritize user engagement over well-being. By focusing on the addictive nature of algorithmic recommendations and the potential harm from repeated exposure to specific content themes, regulators are directly confronting the business model of platforms like TikTok. This investigation sets a precedent for how other platforms with similar recommendation engines will be scrutinized under the DSA, potentially impacting content moderation strategies and algorithmic design across the industry.

The Commission's emphasis on implicit engagement signals versus explicit user preferences in algorithmic recommendations suggests a future regulatory direction. If confirmed, this could force platforms to redesign their algorithms to be more transparent and responsive to user-stated desires, rather than solely optimizing for watch time and interaction. This shift could affect user retention metrics and advertising revenue models that rely on maximizing engagement, prompting a broader industry re-evaluation of how content is surfaced and what metrics are prioritized.