Sunday, June 7
AI Power, Politics, and Pushback
The Trump administration is reportedly in talks to take an equity stake in OpenAI (opens in a new tab), a move that would mark an unprecedented entanglement between the federal government and a private AI company. That political intimacy is already showing cracks: Sriram Krishnan (opens in a new tab), the administration's senior AI policy advisor at the White House, is stepping down from the role he's held since early 2025. Together, these developments signal that Washington's experiment with hands-on AI governance is entering a more volatile phase, where the line between regulatory oversight and ownership is blurring fast.
Illinois has joined Ohio in ordering a pause on data center tax credits (opens in a new tab), a sign that state legislatures are growing skeptical of subsidizing AI infrastructure that delivers uncertain local benefit. The political temperature around these projects is rising at the community level too — the mayor of Shelbyville, Indiana (opens in a new tab), drew national attention after dismissing residents who oppose a planned data center by saying only people in 'shitty houses' would object. Meanwhile, OpenAI is rolling out Lockdown Mode (opens in a new tab), a new security feature designed to guard sensitive enterprise data from prompt injection attacks, suggesting the industry knows its expanding footprint into critical infrastructure demands better safeguards before regulators impose them.
Bots now outnumber humans online (opens in a new tab), and the infrastructure of the internet — built for person-to-person communication — was never designed to handle that ratio, creating systemic pressure on everything from ad markets to content moderation. Meta is accelerating the problem (opens in a new tab) from a different angle, having built an AI-generated clickbait news feed inside its own app, flooding users with synthetic content that mimics journalism without any of its accountability. Young people are registering the cumulative effect: surveys show a growing cohort describing AI as more harmful than helpful (opens in a new tab), a sentiment shift that matters enormously for companies betting their next decade on consumer AI adoption.