Senate Panel Advances Kids Online Safety Act as House-Senate Clash Looms

KOSA Clears Senate Committee as Blackburn and Blumenthal Reject House's Weaker Bill

The Senate Commerce Committee voted on Wednesday to advance the Kids Online Safety Act (KOSA), a bill that would force social media platforms to use “reasonable care” in designing features to prevent a sweeping list of harms to minors under 17—including anxiety, depression, eating disorders, substance use disorders, suicidal behaviors, and “addiction-like” behavior. The legislation targets specific design elements such as push alerts, appearance-altering filters, and videos that play automatically, placing a legal duty on platforms to mitigate those harms through how users interact, not just what they see.

The committee’s move immediately opened a rift with the House of Representatives. Senators Marsha Blackburn (R-Tenn.) and Richard Blumenthal (D-Conn.), the bill’s lead sponsors, denounced a competing House version that omits the broad reasonable-care mandate and instead requires only “reasonable” policies regarding threats of violence, sex abuse, drug use, and financial fraud. “The Senate cannot and will not accept the version that the House passed earlier,” Blackburn said after the vote, calling it “toothless” and a “pale imitation of big tech accountability.”

Opposition to the Senate bill grew louder even before the committee hearing. The Electronic Frontier Foundation (EFF) warned that the measure would prompt platforms to censor lawful speech out of fear of liability. When “someone later claims contributed to harms like anxiety, eating disorders, or substance use,” the EFF wrote, “the safest response is to remove lawful speech or shut down forums discussing those topics altogether.” Separately, the tech industry group Computer & Communications Industry Association (CCIA) said the bill’s application to any platform “reasonably likely to be used” by minors would incentivize companies to verify the ages of all users, collecting sensitive personal data from both minors and adults.

With the committee vote complete, the bill heads to the full Senate, where the real battle will be reconciling it with the House-passed language. If no compromise emerges, the legislation could stall, leaving the current patchwork of state-level child-safety laws—and the underlying design practices of social platforms—untouched.

Inside the Senate Bill's 'Reasonable Care' Standard and Its Collision with Free Speech and Privacy

The Senate’s Duty-of-Care Model vs. the House’s Narrower Standard

The Senate version of KOSA departs sharply from the House bill by creating a broad legal duty for platforms to “prevent and mitigate” a catalogue of psychological and behavioral harms. It does not merely regulate content but rewires the product decisions behind engagement: autoplay, notifications, and augmented-reality filters. This duty-of-care approach raises the stakes from content-moderation policies to product architecture itself. Platforms would face potential enforcement from the Federal Trade Commission and lawsuits from state attorneys general if they cannot show they used reasonable care in these design features—an ambiguous legal standard that will likely be tested in court.

The House bill, by contrast, limits its scope to four categories of palpable harm—violence, sex abuse, drug use, and fraud—and only asks for “reasonable” policies, not a design-level obligation. That makes it easier for tech companies to comply but far less interventionist. The Senate sponsors’ outright rejection of the House language signals a fundamental disagreement about the proper reach of federal platform regulation. If the Senate insists on its version, the next phase becomes a conference negotiation where the House may dig in, given members’ wariness of overbreadth and free-speech challenges.

Free Speech and Age Verification: Two Critical Friction Points

The EFF’s critique gets to the heart of the First Amendment problem: by tying liability to harms like anxiety or eating disorders, the bill essentially makes platforms responsible for how third-party content might affect a minor’s mental state. That could lead to aggressive over-censorship—removing lawful discussions of mental health, body image, or addiction because they might later be cited as contributing factors to a listed harm. For platforms that rely on ad revenue from broad, engaged audiences, this chilling effect could narrow the range of permissible speech and reduce the diversity of content that keeps users online.

The CCIA’s privacy warning is equally consequential. Any platform that is “reasonably likely to be used” by a minor—a vague standard—would be pushed toward verifying the age of every user. That means collecting government-issued identification, facial scans, or financial credentials, creating a massive new repository of sensitive data. For a company like Meta or TikTok, storing that data securely adds significant compliance cost and creates a honeypot for hackers, while smaller platforms might lack the resources to do it at all. Both objections make the Senate bill a target for legal challenges on multiple fronts if enacted.

What This Means for the Ad-Supported Social Media Economy

The features the bill directly targets—push notifications, autoplay, and appearance-altering filters—are among the most powerful tools for driving engagement and time-on-platform, which in turn fuel the ad revenue crucial to social media companies’ business models. Redesigning these features to avoid liability under a reasonable-care obligation could reduce session lengths, depress ad inventory, and force platforms to rethink their core growth loops. At the same time, a platform that can credibly present itself as “safer” for kids may attract brand advertisers nervous about appearing next to harmful content, creating a potential competitive differentiator.

However, enforcement remains unclear. The FTC would need to define “reasonable care” in a design context, a process that could take years and generate litigation over due process and vagueness. In the interim, tech companies would face a compliance vacuum, with the threat of enforcement hanging over every product update. That uncertainty alone could freeze innovation in features aimed at younger users, pushing age-gating or even separate, sanitized versions of apps for minors—a path YouTube’s separate kids app has already trodden.

Next Moves for Tech Platforms Facing a Likely Senate-House Showdown

• Map product features against the Senate bill’s harm list. The bill explicitly names features like push alerts, autoplay videos, and appearance-altering filters. Audit these for potential links to anxiety, depression, eating disorders, or addiction-like behavior. Document current design choices and what design alternatives would cost, so you can argue “reasonable care” was exercised if challenged.
• Model age‑verification scenarios now. Because liability attaches to platforms “reasonably likely to be used” by minors, begin testing age‑estimation or verification methods that minimize data storage. Prepare for the possibility that all U.S. users may need to prove their age, and evaluate the infrastructure, cost, and privacy compliance required. The CCIA’s specific warning about incentivized age checks is a signpost of what enforcement could demand.
• Watch the Blackburn‑Blumenthal vs. House dynamic. With the lead senators publicly rejecting the House version, the fate of any final bill hinges on a conference negotiation. Companies with active congressional relations should engage now to shape a compromise that avoids the most onerous speech‑policing and privacy mandates without gutting the child‑safety core—waiting until a bill reaches the president’s desk is too late.
• Reinforce First Amendment defenses in content‑moderation frameworks. Legal challenges on free‑speech grounds are almost certain if a duty‑of‑care approach becomes law. Begin preparing internal analyses of how liability for emotional harms could intersect with Section 230 jurisprudence, and document current moderation policies so you can show they were not designed to censor lawful speech. This is not about litigation forecasting—it’s about being able to demonstrate a rights‑respecting process from the start.