The House passed the Kids Internet and Digital Safety Act (H.R. 7757) on June 29, 2026, by a bipartisan 267-117 vote, folding a revised Kids Online Safety Act, a COPPA update, an AI-chatbot safeguard regime, restrictions on minors' private and disappearing messages, and mandatory age verification for pornography sites into one 14-bill package. Committee leaders Brett Guthrie, R-Ky., and Frank Pallone, D-N.J., who co-sponsored the bill, called it a bipartisan answer to "design features, default settings, and kids' privacy" (IAPP). It now goes to a Senate where the two parties disagree on how far platform liability should reach.
The steelman
The case for the bill's core impulse is genuinely strong. Congress has spent years documenting how algorithmic feeds, autoplay, and infinite scroll interact with adolescent psychology in ways platforms' own internal research has flagged. AI chatbots marketed as companions raise a distinct problem: a system that can simulate empathy at scale, with no clinician, no license, and no duty of care, talking to a 13-year-old at 2 a.m. The KIDS Act's chatbot provisions — requiring bots to disclose they aren't human, share mental-health resources, and prompt breaks — are a proportionate response to that specific gap, not the sweeping restructuring EFF and other critics warn about elsewhere in the bill.
The age-verification piece for adult sites also has a real constitutional foundation, not an invented one. In Free Speech Coalition, Inc. v. Paxton, the Supreme Court ruled 6-3 on June 27, 2025, that Texas's age-verification law for sites where more than a third of content is sexually explicit passed intermediate scrutiny, holding that it "only incidentally burden[ed] the protected speech of adults" (Justice Thomas, writing for the majority, joined by Roberts, Alito, Gorsuch, Kavanaugh, and Barrett; Wikipedia summary of the opinion). More than 20 states have since enacted or enforced similar laws. The KIDS Act's SCREEN Act component tracks that precedent closely: it gives covered adult sites one year to comply, and — per a Crowell & Moring analysis of the enacted text — "does not prescribe a particular verification technology or require users to submit government issued identification" (Crowell & Moring). That is a narrowly tailored provision sitting squarely inside settled law.
Where the bill overshoots
The rest of the package does not stay inside that lane. Paxton blessed age verification for a specific, narrow category — commercial sites principally trading in content obscene to minors — precisely because the incidental burden on adult speech was contained. The KIDS Act's broader youth-privacy provisions apply a "knows or should have known" standard to whether a user is under 13 or between 13 and 16, across general-purpose platforms that host every other kind of speech. The Electronic Frontier Foundation's read of the bill is that this creates enough legal exposure that platforms will build age-verification infrastructure well beyond what the statute technically requires, just to manage liability risk — even though the bill's own disclaimers say verification isn't mandatory (EFF). That is a predictable compliance dynamic, not a hypothetical one: risk-averse general counsel offices routinely over-comply with ambiguous liability standards, and the Paxton Court's tailoring logic doesn't travel to a rule that reaches every user of a mainstream app rather than visitors to a narrowly defined adult site.
The mechanism matters, too. EFF's other concern — that current age-estimation systems "make mistakes when estimating children's ages correctly" and fail disproportionately for people of color, people with disabilities, and trans and nonbinary users — is not a fringe objection; it's the same accuracy critique that dogged facial-analysis tools before regulators forced disclosure of error rates. A federal mandate that pushes platforms toward those tools without accuracy or bias standards imports that problem at national scale.
"[The bill] requires platforms to 'establish, implement, maintain, and enforce' policies addressing broad categories including narcotic drugs, tobacco products, cannabis products, gambling, and alcohol" — language EFF warns could sweep in teens seeking harm-reduction or addiction-recovery information, not just content that promotes those activities.
The messaging provisions raise a related problem. Encryption carve-outs exist for some features, but they don't cover KOSA's requirement that platforms "address" harms occurring inside encrypted conversations — pressure that, in practice, points toward weakening encryption or curtailing the ephemeral-messaging features many teens (and adults) rely on for ordinary privacy, not just wrongdoing.
The Senate fight is about something else entirely
Ironically, the Senate objection likely to determine the bill's fate isn't the First Amendment critique — it's the opposite complaint. Senators Richard Blumenthal, D-Conn., and Marsha Blackburn, R-Tenn., have said the House version is too weak because it drops the Senate KOSA's "duty of care" standard in favor of "reasonable policies and procedures" language. Blumenthal has called the House bill "a blank check to Mark Zuckerberg." That fight — over whether platforms face an affirmative legal duty to redesign features versus a process-based compliance standard — is a legitimate design question. A duty-of-care standard invites courts to second-guess product design choices after the fact, which is a broader and vaguer liability regime than the KIDS Act's current text, not a narrower one.
The proportionate path
Congress doesn't need to choose between an unaccountable internet and an identity-checked one. The porn-site age-verification model that survived Paxton — narrow scope, no ID mandate, minimized data retention — is a workable template precisely because it stayed narrow. Extending "knows or should have known" liability across general platforms, without accuracy standards for the age-estimation tools that liability will summon into existence, trades a real but bounded harm for a diffuse one: a verification layer sitting in front of ordinary speech for millions of adults who were never the target of the bill in the first place.