The push for TikTok's records traces back to a lawsuit filed by the Kentucky Attorney General in October 2024, part of a broader investigation by attorneys general in 14 states (Britt release, October 2024). Portions of the complaint had been poorly redacted, and NPR's reporting on those faultily redacted sections became the basis for a public statement from Senator Katie Britt.

As relayed in that statement, TikTok's own research reportedly found a user "likely to become addicted" to the platform within 35 minutes, and internal findings flagged effects including anxiety and declines in memory, empathy, and analytical thinking (Britt release citing NPR, October 2024). The same filings reportedly cited internal data showing that 95% of smartphone users under 17 open TikTok at least once a month (Britt release citing NPR, October 2024).

None of that amounts to a settled clinical finding. It's an allegation drawn from a redacted complaint and relayed through a senator's statement rather than the unsealed filing itself. What TikTok actually measured to reach a "likely to become addicted" conclusion, over what population and using what definition of addiction, has not been made public.

Blackburn and Blumenthal, who co-authored the Kids Online Safety Act, wrote to CEO Shou Chew in October 2024 requesting all documents, communications, and research TikTok held on minors' safety, asking the company to hand over the same material it had already produced for state attorneys general (Blackburn release, October 2024). Their letter accused TikTok of "knowingly" designing products that foster "destructive addiction" while misleading the public about the platform's safety, an accusation that, absent a court ruling, remains a claim rather than an adjudicated fact (Blackburn release, October 2024).

KGI's report, published in February, put numbers behind those 2024 allegations. Reviewing litigation records, KGI found that between January 2024 and January 2025, only 1.5% of TikTok users turned on screen-time break reminders, and between 0.7% and 1.8% enabled sleep reminders (KGI, February 2026).

TikTok isn't an outlier here. Meta's internal projections accurately predicted that 99% of teens would ignore its optional "Take a Break" prompts on Instagram and Facebook (KGI, February 2026). As of last year, only 0.15% of minors had enrolled in Facebook's parental supervision tools, and just 0.0038% of Instagram users had adopted parental controls (KGI, February 2026).

Low adoption doesn't mean a tool fails the small share of users who turn it on. According to KGI, it does mean a setting almost nobody activates cannot function as protection for the wider teen population.

KGI ties these adoption numbers to a wider set of concerns raised in both EU risk assessments and US litigation: compulsive use, sleep deprivation, self-harm, eating disorders, and other mental and physical health effects on minors (KGI, February 2026). TikTok's internal research specifically documented compulsive-use patterns and evaluated its safety tools using detailed behavioral data, the same review found (KGI, February 2026).

Meta's internal records show a starker example of what can slip through when platform design goes unchecked. Instagram's "Accounts You Might Follow" feature recommended adult accounts tied to grooming behavior to nearly 2 million minors over a three-month period in 2023 (KGI, February 2026). KGI cites the example to make a point that extends across both companies: platforms hold detailed, granular knowledge of how their products affect young users, but that knowledge rarely surfaces in what they disclose to regulators.

Under the EU's Digital Services Act, platforms are required to identify and mitigate risks to minors. The risk assessments filed so far have mostly offered high-level descriptions of policies and tools rather than data showing whether those tools reduce harm, according to KGI (KGI, February 2026).

Britt called the 2024 disclosures "disgusting" and said they were further evidence that "Big Tech prioritizes profits over people" (Britt release, October 2024). She linked the moment to three pieces of legislation: KOSA, which passed the Senate in 2024 and which Blackburn and Blumenthal describe as requiring "accountability, safeguards, and transparency" from platforms; the Stop the Scroll Act, which would mandate mental-health warning labels on social media; and the Kids Off Social Media Act, introduced in 2024 with Senators Schatz, Cruz, and Murphy (Britt release, October 2024).

The material reviewed for this article does not include TikTok's response to that October 2024 request, nor does it show whether the company has responded to KGI's February report. Neither Blackburn's nor Blumenthal's office has published TikTok's reply, if one exists, in the record available here. What's documented is a demand for records in 2024 and an independent analysis, published roughly sixteen months later, that gives that demand more specific evidence behind it. What's missing from that record is TikTok's side of the story.

The best-supported claim in this record is narrower than "TikTok ran an experiment on kids." Litigation records reviewed by KGI describe an internal ceiling that limited how much new screen-time tools could reduce usage, alongside separate adoption data showing those tools went largely unused. That's a documented design choice, not proof of harm to any individual teenager, and the record doesn't establish that the guardrail caused a specific mental-health outcome in a specific user.

What remains open, per KGI's analysis, is whether disclosure rules alone push platforms toward defaults that actually work, tools that activate automatically for minors, algorithmic changes, or independent testing of whether a safety feature reduces harm, rather than a setting most users never find. TikTok's own account of the 5% figure and the adoption numbers isn't part of the public record reviewed here. Until it is, the question sits exactly where the 2024 documents left it.