The White House and Congress are reportedly working on a deal to trade free expression online for free markets. Last Monday, multiple high-ranking officials from the Trump administration held meetings to gauge support for a package of kids’ online safety and digital likeness rules in exchange for federal preemption of certain state and local AI regulations. The offer would be tempting to the administration, which has repeatedly tried, and failed, to secure relief from onerous or duplicative state AI rules since last June. The deal is also a rare opening for lawmakers whose kids’ safety bills have languished in the Senate. But, a preemption won on these terms would be a Pyrrhic victory, paid for with Americans’ freedom to speak online. The administration should walk away.
Senator Marsha Blackburn (R-TN) is spearheading conversations with the White House to determine what makes it into the final deal. As of now, the package is rumored to include Blackburn’s Kids Online Safety Act (KOSA), the App Store Accountability Act (ASAA), and the NO FAKES Act. In return, the administration would receive what Blackburn’s office has described as “subject-matter” preemption — only barring states from regulating the topics covered in the package rather than AI regulations broadly. The cost of preemption may continue to grow, as the kids’ safety organizations present at Monday’s meeting are also pushing to add Senator Josh Hawley’s (R-MO) GUARD Act.
Though new information about the talks has trickled out in recent days, the substantive details of the arrangement — particularly how expansive or narrow federal preemption of state AI laws will be — remain unknown to the public. But while the package may still change in the coming weeks, these proposals would carry dire implications for Americans’ privacy and First Amendment rights if enacted.
KOSA
The act would require online platforms to “exercise reasonable care in the creation and implementation of any design feature” to prevent and mitigate harms to minors, including anxiety, depression, or eating disorders. To meet the subjective standard of providing “reasonable care” in “design features,” platforms are likely to suppress, demote, or age-gate protected speech.
The Federal Trade Commission (FTC) is granted expansive authority to determine how platforms should function and what content should be throttled. Since the design features that cause anxiety or compulsive usage are left undefined, the FTC can limit speech the government identifies as politically unfavorable in violation of platforms and users’ First Amendment rights.
ASAA
The App Store Accountability Act would require app stores to verify the age of every single user and obtain parental consent before a minor can download an application or install a “major” update. The mandate draws no distinction based on an app’s content or risk profile, meaning that apps for gambling, checking the weather, and reading the Bible are all subject to the same verification regime. Millions of benign apps would be barricaded behind age screenings, forcing every adult to prove their identity to reach lawful content. As we have warned in the past, conducting age verification at the app store level instead of the platform level does not change its fundamental flaws; it simply relocates where the harms occur.
NO FAKES Act
The NO FAKES Act would create a new federal property right in each person’s voice and likeness, focused on mitigating the potential harms of AI-generated replicas. But in doing so, the bill would impose a notice-and-takedown regime for digital platforms that would be incentivized to delete first and ask questions later.
Users would bear the legal burden and expense to restore content erroneously taken down — such as memes, satire, or similar forms of protected speech — chilling free expression online.
GUARD Act
We recently noted that the GUARD Act “would explicitly require every user of an ‘AI companion chatbot’ to verify their age using sensitive information linked to their real-world identity, like a government-issued ID or personal banking data.” Adult users unwilling to sacrifice their privacy and online anonymity by succumbing to invasive age verification would be barred from using AI tools. Similar age verification laws have failed to overcome constitutional scrutiny in the past, making the GUARD Act unlikely to go into effect and ultimately failing to protect younger users.
Conclusion
Keeping digital spaces safe and freeing American innovators from a thicket of state AI laws are both incredibly important goals. That is precisely why they should never be put at odds with one another. If these bills can stand on their own merits, they should be brought forward independently rather than being strapped to the administration’s priorities. Unfortunately, the current negotiations force policymakers to divert from what’s best to protect kids online, instead reaching for a grab-bag of proposals that would erode user privacy and free expression online.
Even without new legislation, families already have a growing array of parental controls and safety tools at their disposal. Companies including Apple, Google, Meta, and OpenAI have expanded their suites of parental tools and safeguards for families to choose from. Parents are far better positioned to determine what best fits their child’s needs than policymakers in Washington. Rather than build on the options families already have, the bargain taking place would impose real and immediate cost on how Americans can express themselves online for a narrow, still undefined preemption of state AI laws. Congress and the administration should refuse it.
