US senators question TikTok over safety feature experiment
Senators seek TikTok explanation
US Senators Marsha Blackburn and Richard Blumenthal have asked TikTok to explain an experiment that reportedly withheld an algorithmic safety feature from 15 million US users. Bloomberg reported that the company disabled the safeguard for 10% of its US audience to create a control group.
The senators sent their letter to TikTok CEO Shou Chew and Adam Presser, CEO of the company’s US business. They described the decision to conduct the experiment as “depraved” and set September 1 as the deadline for a response.
Test examined engagement effects
The algorithmic safeguard was designed to prevent users from being overwhelmed by harmful content. TikTok reportedly wanted to establish whether the measure made the service less engaging, and therefore withheld it from the control group.
The reported group included 16-year-old Chase Nasca. His algorithm served him thousands of videos concerning sadness, suicide and loneliness before he died by suicide. The senators cited the case while seeking details of the experiment and its operation.
Questions focus on accountability and minors
Blackburn, a Republican from Tennessee, and Blumenthal, a Democrat from Connecticut, submitted 13 questions to the company. They are seeking the names of every employee who knew about the experiment and an explanation of why TikTok permitted minors to be included in the control group.
They also requested a list of every algorithmic experiment carried out in the United States in which TikTok disabled or delayed a safety feature. The request would extend scrutiny beyond the single reported test to the company’s wider approach to product experimentation.
Implications for platform testing
TikTok did not immediately respond to a request for comment from TechCrunch. The company now faces a defined congressional deadline to address the reported test, the people involved and the treatment of younger users.
For businesses operating digital products, the case underlines the need to define safeguards, participant eligibility and internal accountability before testing changes that can affect exposure to harmful content. Engagement measurement does not remove the need to document how safety protections are handled during experiments.