Addictive Design Under Scrutiny: The EU, Iowa and California

Three recent developments concern addictive design: preliminary findings in the EU, a court ruling in Iowa, and a legal challenge to a California law.

European Commission: the “rabbit hole” effect

The European Commission has preliminarily found TikTok in breach of the Digital Services Act for the so-called rabbit hole effect of its recommender system. TikTok’s “For You” feed shows users a continuous stream of personalised videos, including videos related to mental health struggles and body image.

According to the Commission, the more a user interacts with a type of content, the more the algorithm recommends similar content. It says this design can be harmful when the repeated exposure involves content about sadness, isolation or body image, with possible effects on mental health, especially for vulnerable adults and minors.

The preliminary findings cover two areas. On risk assessment under Article 34, the Commission preliminarily found that TikTok did not adequately assess how its algorithm may lead users into rabbit holes. It says the algorithm treats repeated watching or liking of content, such as weight-loss or body image content, as a positive sign of interest and recommends more of the same.

On mitigation under Article 35, the Commission preliminarily found that TikTok failed to implement proportionate, reasonable and effective mitigation measures, and that the measures it has in place are not effective. In its view, measures to recommend diversified content do not sufficiently reduce how often similar content is recommended, and the control tools do not effectively let users control what they see. Using the “not interested” button, for example, may not be enough to change the feed, and the algorithm prioritises viewing time over a user’s choice to see less of certain content.

At this stage, the Commission considers that TikTok needs to change the design of the feed. It gives examples: relying more on users’ explicit preferences, disabling viewing time as an indicator for recommendations, increasing the accessibility and usability of user controls, and regularly testing the algorithm.

The findings are part of formal proceedings on TikTok’s compliance with the DSA. In the same proceedings, the Commission adopted preliminary findings on addictive design in February 2026 and on the protection of minors in July 2026. It also adopted preliminary findings on researcher access to public data in October 2025, and closed the advertising transparency strand through binding commitments in December 2025.

Iowa: claims against TikTok proceed

In 2024, the State of Iowa sought a permanent injunction under the Iowa Consumer Fraud Act over statements by TikTok about the frequency and severity of content such as alcohol, tobacco and drug content, sexual content, nudity and profanity, its age-rating statements in app stores, and its statements about Restricted Mode.

In April 2026, Attorney General Brenna Bird amended the suit to add two claims. The first asserts that TikTok uses dangerous, addictive features to cause children to become hooked on the app. The second asserts that TikTok misrepresented its data collection practices and ties to China, including that user data was still being sent to ByteDance and so accessible to Chinese officials.

TikTok sought to dismiss the two new counts. An Iowa district court judge partially denied the motion, and the suit may proceed with both. The ruling also addressed Section 230 of the Communications Decency Act, which provides that no provider or user of an interactive computer service shall be treated as the publisher or speaker of information provided by another. According to the court, Section 230 is not a general immunity from liability deriving from third-party content. The State’s claim does not seek to hold TikTok liable for publishing information, but for designing a platform that allegedly encourages or causes addictive or compulsive use. The judge did not discuss product liability law but described the claim as treating TikTok “as a products manufacturer” rather than a publisher.

TikTok also argued that the suit would intrude on its expressive activity, namely the design features it uses to present content. The court said the suit does not target expressive activity. It referred to a California court’s reasoning that personalised feeds rely on predictive algorithms that use users’ past watch history and other data to suggest content likely to keep them engaged, which means platforms are not making decisions about what content is interesting.

California: a challenge to AB 1709

Google, Meta and TikTok have each filed a separate federal lawsuit to block California’s Assembly Bill 1709. The law prohibits platforms from offering users under 16 “psychologically exploitative features” that could lead to compulsive use, such as personal recommendations and autoplay. Governor Gavin Newsom signed it on September 10, and enforcement is set to begin on January 1, 2027, unless a federal court issues an injunction.

The companies make two main arguments. The first is that Section 230 protects the organisation of content by algorithm from state liability and design mandates. The second is that algorithmic feed curation is protected speech under the First Amendment.

Each company also commented. Google spokesperson Jose Castaneda said that parents are best positioned to make the right decisions for their families and that overly broad laws like this break the systems built to help children find age-appropriate videos. Meta said California cannot override its editorial judgments about how to curate and display billions of posts, while leaving children free to spend unlimited time on other platforms offering similar services. TikTok said protected speech “cannot be suppressed solely to protect the young from ideas or images that a legislative body thinks unsuitable.”

Fairpatterns detects recommender feeds, autoplay and other addictive design features automatically across platforms and at scale.