June 29 Tech and Democracy Roundup: The Congressional Standoff on Kids’ Online Safety
Why the House and Senate can’t agree.
Welcome back to your Tech and Democracy Roundup. Happy early 250th birthday to America.
I want to take you back five years, to one of the biggest whistleblower moments of the social media era. It’s October 2021. Former Facebook product manager Frances Haugen discloses thousands of pages of internal documents – known as the “Facebook Papers” – to the SEC, revealing that the company knew its products were harming teenage girls and amplifying inflammatory content. (This fall, a companion film to The Social Network tells Haugen’s story. Early reviews seem lukewarm. But Mikey Madison as Haugen and Jeremy Strong as Zuckerberg? Worth the price of admission).
The Facebook Papers catalyzed a new wave of advocacy to hold social media companies accountable, culminating in the passage of two landmark Senate bills in July 2024: the Kids Online Safety Act (KOSA) and COPPA 2.0. But neither ever received a vote in the House, where Republican leadership has instead produced competing social media legislation. Their latest package, unveiled last week, includes a pared-back version of KOSA, along with a hodgepodge of provisions on kids’ safety and data privacy. The bill is set for a vote on the House Floor today (June 29). It too appears ill-fated in the Senate.
Why can’t the two chambers get on the same page?
“Kids’ online privacy and safety legislation attracts broad ‘save the children’ commitments that tend to break down on critical specifics that companies don’t like,” Lindsey Barrett, an assistant professor of law at Drexel University’s Kline School of Law, told me over email.
When it comes to the impasse between the House and Senate, two sticking points in particular stand out: duty of care and state preemption. Specifically, should social media companies be legally required to “exercise reasonable care” to prevent harms to minors? (Senate says yes; House says no.) Should states have the right to pass their own social media legislation? (Senate says yes; House says no.)
Responding to the House proposal last week, Sens. Richard Blumenthal (D-CT) and Marsha Blackburn (R-TN), co-authors of the Senate’s KOSA bill, called the package dead on arrival, citing the bill’s removal of the duty of care provision. Opponents of the provision argue it could force platforms to over-censor. Senator Cantwell (D-WA) added that the Senate “is not interested in having state cases preempted”, referring to the preemption language in the House bill.
The House and Senate appear deadlocked for now. But continued gridlock may not be the worst outcome.
“The prospect of not passing something fast enough doesn’t worry me,” Barrett wrote. “I’m much more concerned about performative bills that heap busywork on the FTC and celebrate cosmetic disclosure requirements being embraced as meaningful reform.”
Now, onto the other big headlines.
White House
The Trump administration asked OpenAI to limit the release of its latest model, GPT-5.6, to a small set of government-approved partners before a wider public release, a move that echoes the export control directive on Anthropic earlier this month. Critics of the move (including tech executives who have been supportive of the President) argue it is the latest example of the administration’s ad hoc approach to AI model oversight.
The administration is pressing Meta to submit its AI models for government safety reviews. It is the only major AI company that has yet to agree to the voluntary process.
The administration is providing a $17.5 billion loan for 10 new large nuclear reactors to meet power demand from AI data centers.
Congress
Several new AI bills were introduced:
Sen. Bernie Sanders (I-Vt.) introduced the American AI Sovereign Wealth Fund Act, which would levy a one-time 50% tax on the stock of the largest AI companies to create an estimated $7 trillion sovereign wealth fund.
Sens. Brian Schatz (D-HI) and John Curtis (R-UT) introduced the AI Labeling Act, legislation to require AI providers to visibly label AI-generated content.
Rep. Nathaniel Moran (R-TX) introduced the AI Incident Reporting Act, legislation to require AI developers to report critical safety incidents to the Secretary of Commerce.
Rep. Alexandria Ocasio-Cortez (D-NY) introduced the AI Data Center Moratorium Act, which would temporarily prohibit the construction of new data centers until “Congress passes more comprehensive AI legislation”.
The House Science Committee advanced 10 bipartisan AI bills out of committee, including measures to establish a national AI research resource, create standards for detecting AI-generated content, and improve transparency around AI model development.
States
California Gov. Gavin Newsom signed an executive order directing state agencies to study AI’s workforce impacts and explore how to share AI’s economic benefits more broadly – including launching a public dashboard tracking AI-related job losses.
A bipartisan coalition of over 200 state lawmakers signed a letter urging Congress to oppose the Great American AI Act, a proposal released earlier this month by Reps. Jay Obernolte (R-CA) and Lori Trahan (D-MA) that would create a federal framework for AI development, preempting state AI laws for three years.
Courts
The Department of Justice intervened in support of xAI in a lawsuit filed by the NAACP alleging the company is endangering public health by running unpermitted natural gas turbines at its Colossus 2 data center in Southaven, Mississippi. In its filing, the DOJ argued that attempts to stop xAI from running the turbines “threatens American national, economic, and energy security”.
A nationwide coalition of publishers representing nearly 400 newspapers sued OpenAI and Microsoft for scraping their content without permission or compensation, warning that unchecked AI copyright infringement could be a “death knell for local journalism”.
Industry
Google became the latest major AI company to publish a policy framework. Their white paper proposes a “pragmatic middle path” on AI governance, most notably an independent (industry-funded) regulatory body to oversee frontier AI labs.
Meta announced plans to replace 90% of its remaining human content moderators with AI by the end of 2026 – a significant retreat from human oversight.
Civil society
Former Commerce Secretary Gina Raimondo and Indiana’s former Republican governor Eric Holcomb announced RAISE US, a nonpartisan initiative to help workers adapt in the age of AI. The effort has received over $500 million in funding, including from the OpenAI Foundation and Anthropic.
Raimondo is also co-chairing, alongside former House Speaker Paul Ryan, a new bipartisan commission launched by AEI and the Urban Institute on AI and the workforce.
A new report by immigration advocacy organizations found that ICE spending on AI-powered surveillance grew from $310 million in 2025 to a record $513 million in 2026, fueled by new contracts with Palantir and Anduril.
A new report from Yael Eisenstat (Cybersafety Research Center) and Justin Hendrix (Tech Policy Press) offers nine recommendations for platforms to mitigate political violence.
A new report from Americans for Responsible Innovation calls for national safety benchmarks for AI chatbots used by minors.