Technology

TikTok gave some teens a placebo feed reset that changed nothing, New York says

Adrian Kessler
Add us on Google

When thousands of TikTok users, teenagers and children among them, pressed the app’s button to wipe their recommendations and start over, nothing happened, according to newly unsealed filings in New York’s lawsuit against the company. They were shown a message saying the reset had worked. Their For You feed kept serving the same kind of videos.

The tool at the centre of the claim, Algo Refresh, is one of the few exits TikTok offers someone whose feed has drifted somewhere they no longer want to be. It is meant to make the For You page behave as if the account were brand new. New York alleges that some of the people who reached for it were placed in a test group and handed a version that did nothing, while users outside the experiment got one that worked.

The complaint says TikTok then compared the two groups on time spent in the app and on advertising revenue. It also quotes an internal objection. A product manager on TikTok’s Digital Well Being team wrote in a chat that “This feature is about algorithm transparency and control,” adding: “A placebo group completely conflicts with that.” New York further claims that even the working reset was only temporary.

TikTok rejects that account. A spokesperson told Reuters that “we regularly test new products and features to validate the experience and understand how they work in the real world,” and said the lawsuit gives a false picture of how Algo Refresh works. In a court filing, the company acknowledged that its tests “may involve a portion of platform traffic that continued to receive the original experience,” but denied that this left those users any less safe.

There are real limits to what is known. These are allegations in an amended complaint, not findings by a court, and the passages made public so far give no figure for how many people landed in the placebo group, how long the test ran or what it found. Testing a change against a control group is standard practice across the tech industry. The dispute is whether a safety tool used by minors should ever be one of the things tested that way. The case concerns New York, and the filings reported so far do not say where the tested users lived.

For anyone relying on the button today, the practical point is narrower. A reset is not proof that the algorithm has forgotten you. TikTok’s keyword filters, its restricted content setting and the option to hide specific creators are separate controls, and the reset leaves them in place.

The tests described in the filing took place in 2023, the same year TikTok rolled out the option to refresh the For You feed. New York Attorney General Letitia James first sued the company in October 2024. The placebo claims come from an amended complaint filed under seal at the end of August and ordered unsealed by a judge, as Reuters reported on October 7.

It is not the only experiment in the case. Bloomberg reported in August that TikTok withheld a separate safety update from about 10% of its US users, roughly 15 million people, and that a 16-year-old New Yorker in that group died by suicide in 2022. No court has linked the death to the test.

More than two dozen US states have sued TikTok, accusing it of designing the app to hook children and of misleading families about its safety tools. Alabama, whose case was the first set for trial, settled in September for up to $300 million. No trial date for New York’s case has been made public.

Tags: , , , , ,

Add us on Google

Discussion

There are 0 comments.