TikTok had the fix. By early 2022, the company had developed “filter bubble prevention” — algorithmic changes designed to break up repetitive clusters of suicide, self-harm, and depression content on the For You page. It rolled the safer system out to roughly 90% of U.S. users.
Then it left approximately 15 million people on the old algorithm as a control group, according to a confidential 14-page internal report obtained through litigation discovery and reported by Bloomberg Businessweek. The stated goal: measure the impact on daily active users and “core metrics.”
Chase Nasca, a 16-year-old Long Island soccer player, was randomly assigned to that control group on January 25, 2022. He died by suicide weeks later.
What the Algorithm Did to Chase
TikTok’s own internal review found his account “fell into a filter bubble” of suicide content — because its safety controls were switched off “by design.”
After Bloomberg’s inquiries prompted an internal review in March 2023, TikTok’s teams reconstructed Chase’s watch history: 7,563 videos in his final two weeks. About 73% carried themes of hopelessness, loneliness, and suicide survivor testimonials — content that technically didn’t violate platform rules. Nearly 10% did violate TikTok’s own community guidelines against promoting suicide and self-harm.
The day before he died, Chase spent more than five hours on the app, searching for posts about suicide and trains. The last video he watched was of someone crying.
The internal document’s conclusion is stark: his account received a “repetitive amount of content surrounding Suicide and Self Harm and depressive topics, particularly leading up to the date of his suicide.” The report states that TikTok’s safety protections “did not take effect on this user by design.” That phrase — by design — sits like a confession written in corporate language. TikTok had already built the tool to prevent exactly this spiral.
What TikTok Told Congress – and What It Didn’t
Two weeks after completing the internal review, CEO Shou Zi Chew testified that teen safety was a “core priority” — and the backtest went entirely unmentioned.
In written follow-up answers submitted May 4, 2023, TikTok told Congress it “does not believe that use of the platform has caused users to commit suicide.” When asked to list all cases where A/B testing had resulted in harm, the company responded only that its Community Guidelines apply during experiments. No specific incidents. No mention of Chase.
A former trust-and-safety team lead — who wasn’t told about the backtest while it ran — later criticized the methodology directly, saying the control group was far too large and should have excluded minors and dangerous content entirely. Even TikTok’s own proposed fix, shrinking control groups to under 1% with tests capped at seven days, would still leave an estimated 1.8 million users unprotected during experiments.
Parents visiting TikTok’s safety page today are reading policies written by the same company that typed “by design” into a document it subsequently fought to keep sealed. Concerns about secretly tracking users across the broader app ecosystem underscore how rarely such internal practices face public scrutiny.
The Rules Are Still Being Written
Chase’s family lost in court, the internal document remains sealed, and TikTok says it has “refined” its methods — without specifying how.
Chase’s wrongful-death suit was dismissed under Section 230 of the Communications Decency Act. The family has appealed. TikTok maintains its testing practices have since improved but has not detailed what changed. The 14-page internal report remains under court seal. The pattern of covert platform design recalls other cases involving a surveillance app built to operate against users without their knowledge.
The company wrote the safety playbook. It chose not to run it. And it is still writing the rules.
