The Senate's Stunning Rebuke of the AI Moratorium

The Tech Oversight Project

The Tech Oversight Project

— 7 min read
The Senate's Stunning Rebuke of the AI Moratorium

Welcome back to The Dispatch from The Tech Oversight Project, your weekly updates on all things tech accountability. Follow us on Twitter at @Tech_Oversight and @techoversight.bsky.social on Bluesky.

🤖🪦 RIP AI MORATORIUM (FOR NOW): In an roller coaster week, the on-going AI moratorium debate came to a conclusion and a stunning defeat for the Big Tech lobby – losing a key vote to strip the moratorium from the budget reconciliation package in a landslide 99-1 vote. Credit where credit is due for Senator Marsha Blackburn, Senator Maria Cantwell, and Senator Ed Markey for introducing the amendment to strike the moratorium from the underlying bill and for aggressively whipping Senators to stand with their constituents and not Big Tech.

It takes a village: For well over a month, state lawmakers, parents, civil rights groups, unions, tech watchdogs, safety advocates, and organizations on both sides of the ideological spectrum have been laying the groundwork to provide opposition to Members of Congress and let lawmakers know the full set of consequences an AI moratorium would have on children, nondiscrimination laws, and our entire economy. Just some of the actions advocates took:

The Takeaway: Sacha Haworth, Executive Director at The Tech Oversight Project, underscored the gravity of the situation, “Last night’s vote was the latest evidence that when tech measures are brought to a vote, Congress will overwhelmingly choose to take on Big Tech’s biggest harms. The tech lobby presented Congress with a false choice: safety or global competitiveness, and the Senate did not take the bait.”

What’s Next: Congress should act swiftly and keep the pressure on Big Tech companies by passing aggressive legislation, like the Kids Online Safety Act, the Open App Markets Act, and the American Innovation and Choice Online Act. All of these bills will help level the playing field, protect vulnerable people, and spur new innovation in tech and AI. Importantly, this won’t be the last time Congress puts forward an AI moratorium, and it is our sincere hope that Members will think twice about the effect it has on their constituents.

PROBLEMATIC TECH GIANTS SEEKING DOD CONTRACTS: As tensions escalate in the Middle East, Big Tech monopolies are sneaking into the Defense Department and a new kind of arms race is taking shape. It’s the collision of two gold rushes: one in AI investment and the other in defense contracting – and now they’re feeding off each other. Tech firms are building systems they’ll also regulate – advising agencies they’re selling to, and embedding their executives into military ranks.

OpenAI just won a $200 million contract to build “frontier AI capabilities” for the Department of Defense. It’s the company’s first time as a prime contractor for the federal government and one that could cement its role not just in AI development, but in how the U.S. military fights, plans, and processes information.

Then you have Peter Thiel’s Palantir, which has pulled in more than a billion-with-a-B dollars in military contracts since 2017. Palantir’s AI-powered platforms are now embedded across defense and intelligence agencies.

Meanwhile, companies including Meta and Oracle are staffing up to compete for military dollars. Meta is partnering with Anduril on AI tools for battlefield deployment and decision-making, while also embedding its executives into the Army’s new “Detachment 201” –  a Palantir-linked program that recruits and commissions tech leaders to work part-time inside the Pentagon.

Elon Musk’s SpaceX is already a big-time defense contractor through Starshield, a classified version of its Starlink satellite program,which the Pentagon is already using for missile tracking and military-grade comms. Anduril, Shield AI, and Scale AI are pushing drone swarms and battlefield automation systems that, until recently, sounded like DARPA fever dreams.

This shift isn’t exactly subtle. Tech companies once avoided military work like it was radioactive. Remember when Google employees revolted over Project Maven? That line’s gone. Defense contracts are a growth market. Investors know it. The Pentagon knows it. So does Big Tech.

So now we have the future of warfare being determined by the very same tech CEOs whom practically the entire country distrusts. What could go wrong?

PUSHBACK: Some lawmakers are raising alarms about how Big Tech AI tools could be used to amass sensitive personal data. For example, ten Members of Congress just sent a letter pressing Palantir’s CEO on reports that he’s helping the government build a centralized database on Americans. Republican Congressman Warren Davidson of Ohio raised concerns, calling the idea a “digital ID" with abuse written all over it.

RULES FOR THEE, NOT FOR ME: Meta’s latest power move is a 15-billion-dollar walking antitrust investigation waiting to happen. Zuckerberg and Co are attempting to skirt the merger and acquisition review process by engaging in an “acqui-hire” – when a company is acquired primarily for the talent of its employees, but in this case, it’s more.

Meta is buying a 49% in Scale AI – significantly over its current valuation, acquiring a board seat, and making a side deal poaching founder Alexandr Wang. Since the acquisition, Meta named Wang to run Meta’s Superintelligence Labs – SOUNDS LIKE SOMETHING THAT DESERVES A MERGER REVIEW.

The problems with monopolies are that they have such a huge cash advantage that they can simply buy their way into new markets and have an advantage any law-abiding company doesn’t have. Remember: Threads overtook X in daily active users by buying their way in. And that’s shaping up to be the case in artificial intelligence.

This a strategic end-run around antitrust rules, and the Department of Justice and the Federal Trade Commission should stand up and take notice. Meta is simply ignoring rules they don’t believe applies to them.

☣️ HAZARD TO YOUR HEALTH: Social media warning labels are moving full steam ahead in states across the country, backed by public health experts, bipartisan lawmakers, and growing public pressure. These bills require platforms to display time-based warnings to young users about the effects of excessive screen time – effectively a surgeon’s warning like on packs of cigarettes.

In California, Assemblymember Rebecca Bauer-Kahan’s AB 56 heads will be heard by the Senate Judiciary Committees this morning after passing the Assembly with broad support – and despite heavy tech industry pushback.

In New York, a similar bill has passed both chambers and now sits on Governor Hochul’s desk. The question is whether she’ll sign it. A5346 would require social media platforms with "predatory features," like infinite scrolling and addictive feeds, push notifications and like counts, to display advisories similar to those found on alcohol and cigarettes.

The bills in New York and California aren’t outliers. They’re part of a national trend. From coast to coast, lawmakers are responding to growing evidence—and growing public demand—for guardrails on social media’s most manipulative features.

  • Minnesota has already enacted a warning label law.
  • Colorado’s law requiring pop-ups for under-18 users will come into effect next year.
  • The Texas House recently approved a similar measure. Other states are watching closely.

Meanwhile, despite public support and mounting evidence about the links between social media use and teen anxiety, depression, and sleep disruption, Big Tech lobbyists are marking the same tired arguments against any and all attempts at regulation – even commonsense ones like warning labels with broad support in the medical community and on both sides of the aisle.

📲 WEAPONIZING THE STREISAND EFFECT: A new report from CheckFirst and Reset Tech, covered by WIRED, reveals the dramatic expansion of “Operation Overload,” a coordinated Russian disinformation campaign that has evolved into a sprawling, AI-powered effort targeting Western media, researchers, and the public.

Since September 2024, the campaign has:

  • Sent nearly 1,000 emails to fact-checkers, seeding false content about Ukraine, Moldova, and other geopolitical topics.
  • Flooded X (formerly Twitter), Bluesky, Telegram, and now TikTok with deepfakes, fake news screenshots, QR codes, and AI-generated imagery.
  • Activated over 11,000 inauthentic accounts, including crypto-themed amplification networks and verified Kremlin influencers.
  • Impersonated over 180 public figures and 180+ organizations including the BBC, USAID, and Reporters Without Borders.
  • Exploited global news cycles and election periods in the U.S., Germany, Poland, and Moldova to sow chaos and distrust.

Since late 2024, nearly 1,000 fake “tips” have been emailed to more than 240 newsrooms and fact-checkers, often linking to fabricated content on Telegram or social media. The campaign uses “content amalgamation”—posting the same false story across multiple formats (fake articles, doctored videos, AI images) to create the illusion of credibility and overload fact-checkers. Falsehoods amplified by influencers are sometimes debunked only after going viral. The report urges platforms and regulators to act faster, especially ahead of global elections.