June 1 Tech and Democracy Roundup: Trump, the Pope, and the Future of Search
What Google's AI search overhaul portends for democratic discourse
It’s Monday, June 1, 2026 — welcome back to your Tech and Democracy Roundup.
Over the past two weeks, some tech policy developments received considerable attention from the mainstream press:
President Trump opted not to sign an executive order that would have given the government early access to new AI models. The 11th-hour decision reportedly came after President Trump fielded calls from Mark Zuckerberg, Elon Musk, and David Sacks. The fallout has exposed factions inside the administration and within the Republican party over how to regulate AI.
The Pope released “Magnifica Humanitas”, his long-awaited encyclical on AI, in which he warns that control over AI and digital infrastructure must not be left to a handful of private actors alone. The Renovator has been covering the Pope’s fascinating influence on the AI discourse for a while now; this is his most important statement yet. Anthropic co-founder Chris Olah stood with the Pope at the encyclical’s presentation.
While other stories went under the radar:
Illinois passed America’s strongest AI safety bill yet. The bill will require OpenAI, Anthropic, and Google DeepMind to have their safety practices audited by a third party. Governor JB Pritzker says he’ll sign it.
OpenAI announced new partnerships ahead of the Midterms — with the Associated Press to provide live election night vote counts, and with Democracy Works to provide accurate voting and registration information.
We’re covering other notable developments below. But first, I’d like to draw your attention to an update to Google Search.
This month, Google unveiled the biggest change to its search box in over 25 years: instead of being directed to top-ranked websites, searchers will now receive an AI-generated summary and be prompted to continue the conversation with Google’s Gemini model. Executives pitched the evolution as a shift from a ‘box that hands you a list of websites to a box that completes tasks for you.’
I raise this because it feels like we are on the precipice of a fundamentally new kind of information landscape — one in which algorithms not only curate the news we see, but AI generates the news for us.
Throughout the 2010’s, social media corroded democratic discourse in well-documented ways — spreading misinformation, fueling polarization, amplifying outrage. But for all its ills, it did provide connection. Our interactions were human-to-human (mostly), albeit mediated by algorithms. If chatbots and AI agents supersede social media as a source of information gathering, what challenges could that portend for democratic discourse?
In the short-term, the effects may be hard to detect. A 2024 Pew study found that just 8% of U.S. adults said Google was their primary source for political coverage — behind Television (35%), news websites (21%) and social media (20%). But as tools like Google and ChatGPT position themselves as convenient news generators, that is likely to change. And we now have fresh data on the risks.
ForumAI, an independent research organization, evaluated 3,136 prompts across four major chatbots and found that 30% of all responses contained at least one verifiable factual error, and nearly a quarter failed a neutrality check.
Ethan Zuckerman, professor at UMass Amherst and director of the UMass Initiative for Digital Public Infrastructure, echoed the neutrality concern, telling The Renovator: “AI systems are inherently biased based on the corpora they are trained on, and those biases can render invisible people who are not well represented in digital media. And when the controllers of AI systems — say Elon Musk with Grok — train their systems to give answers of a particular political valence, these systems do not betray those instructions they’ve been given.”
Another concern is source variety. An MIT study this February found that AI search surfaced fewer independent and niche sources, less variety overall, and significantly more low-credibility sources compared to traditional search.
Mr. Zuckerman also raised a subtler concern that AI summaries erode the deliberative habits democracy depends on: “The AI summary is far more convenient and more comfortable than the list of links. Not only does it provide a singular answer to a query, but it disguises the complexity of differing, and sometimes conflicting, opinions and sources. That makes it an attractive product, but removes some of the useful frictions we need for democracy. Learning to interpolate between a set of possible results for a query is a profoundly important skillset, and one that’s essential for life in a democracy.”
We’ll continue to surface the latest data and research on this topic in future editions. In the meantime, the vulnerabilities of our information environment are a reminder of why verifiable, independent coverage of America’s politics and civic life matters. Consider subscribing to The Renovator below.
Now, onto the other headlines you don’t want to miss.
Congress
Senator Elizabeth Warren (D-Ma.) is arguing for an AI tax. The Senators’ proposal includes overhauling the tax code to incentivize companies to keep human workers.
States
California Governor Gavin Newsom signs an AI executive order focused on worker protections. The order takes a deliberative democracy approach to solicit public feedback on how policymakers should best prepare for the impacts of AI on the future of work.
New York establishes new safeguards to protect kids on social media and online gaming platforms. The state’s Transportation, Economic Development and Environmental Conservation Article bill includes provisions that disable AI chatbot features for kids.
Ride-share drivers in Massachusetts formally unionize. The App Drivers Union became the first organization in the country to be formally certified to represent ride-share drivers. 70,000 workers in Massachusetts now have the power to collectively bargain.
Civil Society
An international team of researchers, journalists, and scholars launches the AI Resist Lab. The public database will catalog efforts to challenge the unchecked expansion of the AI industry.
The Center for Civil Rights and Technology releases AI in the Racial Wealth Gap. The report examines AI’s documented impacts across housing, lending, and employment, and makes the case for interventions to ensure AI doesn’t amplify the racial wealth gap.
The Coalition for Independent Technology Research is suing the Trump administration over the future of online safety. The lawsuit accuses the government of using immigration policy to stifle free speech and tech regulation.
Industry
Anthropic employees pledged to give away billions. There’s a raging debate over where that money will go.
Local
Michigan Township Leader Resigns over Data Center Death Threats. The treasurer of Saline Township, Michigan cited death threats she’d received related to the construction of an Oracle and OpenAI data center.
Texas County passes 1-year data center construction ban. Hill County, southwest of Dallas, passed the state’s first county-level moratorium on data centers.
A retired tech executive helped kill a proposed data center in his Wisconsin village. Now he’s teaching other towns how he did it. (I highly recommend giving this one a read!)
That’s all for now. Until next time. Have a great week!
— Zachey