Data Centers, Dividends, and Dark Money
The tech policy fights you need to know about
It’s Monday, May 4, 2026 — welcome to your Tech and Democracy Bi-Weekly Roundup. My name is Zachey Kliger, and I’m a new staff writer at the Renovator covering tech and democracy. Every other week I’ll be in your inbox with the stories from the world of tech that matter most for democracy, explaining why they matter beyond the headlines. I’ll also track the key policy developments across the White House, Congress, the courts, and civil society.
The tech policy world moves fast. Here’s a sampling of developments from just the last week alone:
The White House appears to be extending an olive branch to Anthropic, whose AI models were blacklisted from Pentagon classified networks earlier this year.
In the Senate, a bill to ban AI companions for minors cleared the Judiciary Committee and will move to the full Senate floor with bipartisan momentum.
At the state level, Maine Gov. Janet Mills (D) vetoed what would have been the nation’s first statewide data center moratorium. And Connecticut became the latest state to pass comprehensive AI regulation.
In industry, Google followed OpenAI and xAI in securing a deal with the Pentagon to provide its AI models, specifically Gemini, for any lawful purpose (a standard Anthropic previously refused). The move sparked an internal revolt at the company.
We’re covering these developments and more below.
But first, I want to return to a question that has been a recurring theme of this roundup: who should govern powerful new AI models — Congress? States? The companies themselves? A still-to-be-formed government agency?
The Trump administration has made clear it doesn’t want states setting those rules. The executive order on AI signed in December created an AI Litigation Task Force with a single mandate: challenge state AI laws.
In April, the task force took its first major action, joining a lawsuit filed by Elon Musk’s xAI against a Colorado algorithmic discrimination law. Other laws in California, Utah, New York, and Illinois are likely next in the DOJ’s crosshairs. (For those interested, Multistate.ai tracks AI legislation across the country in real time.)
This is an alarming development for anyone who cares about curbing AI harm. State laws remain an important, if imperfect, tool for establishing baseline protections while the harder work of federal oversight and cultural adaptation catches up.
For now, state officials appear to be holding their ground: 50 Republican state lawmakers sent a letter to the White House in March urging the administration not to block state AI laws, and attorneys general, lawmakers, and governors have largely resisted pressure to cede their authority. That resolve deserves to be rewarded, but needs to be monitored.
More on this story soon. Now, on to the week’s other big stories.
The Battle over Data Centers
Is the growing, cross-partisan opposition to data centers actually misguided? That’s the provocative argument Holly Buck, an environmental studies professor at the University of Buffalo, makes in a Jacobin op-ed that generated fierce debate on the left last week:
“A moratorium springs from the desire to stop the concentration of wealth, but ironically, it is likely to exacerbate it. Under neoliberal capitalism, industries offshore environmental harms to places with weaker governance, cheaper labor costs, and fewer environmental safeguards. If a better world is the goal, the answer is not merely shifting the geopolitics of AI development but reshaping it.”
I agree that data center moratoria shouldn’t substitute for genuine democratic governance of AI. But Americans deserve a say over who controls the infrastructure of AI, and critical questions — like who foots the bill for utility rate spikes — deserve a public debate.
There’s a small-d democratic silver lining here. Across the country, opposition to data centers is producing genuine civic participation. Ordinary residents are organizing petitions, showing up to zoning meetings, and running for office — in North Carolina, Maryland, and Wisconsin — many for the first time. Whatever you think of the policy merits, that’s local democracy in action.
AI Lobbying, Dark Money, and the Campaign to Shape Public Opinion
The biggest AI companies are spending more than ever to shape AI policy. Anthropic and OpenAI posted their biggest-ever lobbying spends in Q1 2026, per federal lobbying disclosures. In April, Google endorsed over a dozen bipartisan bills on AI’s economic impact and workforce development, and OpenAI published a “social contract” policy blueprint. All of this policy engagement is worth monitoring.
But, while this is happening in clear public view, a more nefarious form of influence peddling is happening behind the scenes. Last week, WIRED reported that Build American AI, a dark-money group tied to Leading the Future — the $125 million super PAC backed by OpenAI, Andreessen Horowitz, and Palantir — has been using social media influencers to amplify pro-AI messaging around the China race narrative.
The line between legitimate policy advocacy and propaganda continues to blur. Voters should be skeptical of where their information about AI is coming from. (This AI Campaign Finance Tracker from Transformer is a useful tool for monitoring political spending from the industry.)
One Answer to AI Job Displacement
It’s arguably the single most critical question facing policymakers in the coming decade: how do we prepare for AI replacing a significant share of human labor?
In April, Alex Bores, a New York assemblymember running for Congress, proposed one potential answer: an AI Dividend — a direct payment program that would kick in if and when AI meaningfully displaces American workers, funded through a tax on AI usage and government equity stakes in major AI companies.
Bores is worth watching. He authored the RAISE Act, an AI transparency law that has become a reference point for other lawmakers, and his congressional platform centers on protecting workers in the age of AI. He has also been the primary target of the OpenAI-backed Leading the Future super PAC — and last week, the New York Times reported that tech billionaire Chris Larsen plans to spend $3.5 million to help Bores fight back against that barrage of attack ads.
Here’s what else we’re following.
In the News
Congress
The Senate gets serious about protecting kids from AI chatbots. The Senate Judiciary Committee unanimously passed Sen. Josh Hawley’s GUARD Act, which would ban AI companies from providing AI companions to minors and require chatbots to disclose their non-human status. A competing bill, the CHATBOT Act from Sens. Ted Cruz and Brian Schatz would instead require AI companies to build “family accounts” giving parents control over how kids use chatbots.
Senators banned from trading on prediction markets. The Senate passed a resolution prohibiting senators from trading on prediction markets — a notable guardrail given how much political intelligence flows through those platforms.
House Republicans propose a national data privacy standard. The SECURE Data Act would establish a single national framework governing how companies collect, share, and sell Americans’ personal data. Industry groups are on board; critics argue the bill limits enforcement to government officials and strips consumers of a private right of action.
Sanders breaks with the pack on China and AI. Sen. Bernie Sanders called on Washington to collaborate with China on AI safety rather than treat development as a race — a notable departure from the bipartisan consensus.
States
Connecticut passed a landmark AI bill. The legislation includes chatbot safeguards for minors, requires disclosure of AI in employment decisions, and establishes an “AI sandbox” program to support innovation. It now heads to Gov. Ned Lamont’s desk, where he is expected to sign.
State Governments are embracing AI — but at wildly different speeds. A new analysis from Code for America finds that Utah, New Jersey, Pennsylvania, North Carolina, Maryland, Texas and Vermont are leading the pack on AI adoption and readiness. West Virginia, Wyoming, Nebraska, Alaska, Florida and Kansas are among the furthest behind.
Courts
The U.S. still considers Anthropic a supply chain risk. The U.S. Court of Appeals for the DC Circuit denied Anthropic’s request to temporarily block the Defense Department’s supply chain risk designation, allowing the ban to remain in effect while the case proceeds.
Massachusetts high court rules against Meta. The Massachusetts Supreme Judicial Court ruled that Meta must face a lawsuit by Attorney General Andrea Joy Campbell alleging the company designed Instagram to addict children — the first ruling by a state high court on whether Section 230 shields platforms from design-based claims.
NAACP sues xAI over Memphis data center. The NAACP filed a federal lawsuit alleging xAI violated the Clean Air Act by operating 27 unpermitted methane gas turbines to power its Colossus 2 data center near majority-Black neighborhoods in Memphis.
Industry
Big Tech cuts workers while spending big on AI. In April, Meta cut 10% of its workforce and Microsoft offered buyouts to 7% of employees, following similar moves by Amazon and Oracle. The reductions come as the same companies pour hundreds of billions into AI infrastructure.
Public Opinion
The AI backlash is intensifying. Separate attacks on Sam Altman’s home and the OpenAI office in April marked a distinct escalation. Meanwhile, new Gallup research on Gen Z finds that the share of young people who feel hopeful about AI has declined since last year, with sentiment becoming increasingly skeptical and negative.
Post of the week (courtesy of Politico’s Digital Future Daily)
Until next time!