According to TechCrunch, Sens. Marsha Blackburn and Richard Blumenthal sent TikTok a letter demanding answers about an experiment that disabled algorithmic protections for roughly 10% of U.S. users. The senators identified one user in that group: 16-year-old Chase Nasca, whose feed reportedly filled with self-harm and suicide content before his death by suicide. TikTok has 13 questions to answer.
This moves beyond platform negligence into deliberate tradeoff testing. TikTok didn't just fail to catch harm. It tested whether reducing harm cost engagement. That's the calculation lawmakers, parents, and regulators will seize on instantly. Congress already views TikTok through a trust lens, and evidence that safety systems were treated as an engagement variable hands critics ammunition heading into 2026.
TikTok's response will determine the political temperature. The company can claim the test was routine product research, but the timing and scale—15 million users, a documented death—make that defense harder. Expect calls for transparency on how many other safety features have been similarly tested.
Tech platforms love the language of testing because it sounds clinical. When the subject is a recommender system steering vulnerable teens toward despair, an A/B test becomes a values test. This one points in a bad direction. The number that sticks is 15 million.
Filed to the Technology desk · Aug 20