Pennsylvania governor demands federal AI safeguards, splitting with tech sector

Josh Shapiro has called for the federal government to step in and regulate artificial intelligence, arguing that developers cannot simply be left to police themselves.

Speaking at the AI Horizons Pittsburgh Summit, the Pennsylvania governor tried to strike a middle path: pushing for fast-paced domestic development while subjecting "frontier models"—the latest, unreleased versions of AI—to strict government scrutiny. He compared the necessary oversight to how the state manages pharmaceuticals and aviation.

"We need the federal government to establish the rules of the road," Shapiro said. "We can’t kick the can down the road anymore."

But inside the Bakery Square complex, where the two-day summit was held, his message was not universally welcomed. Several tech executives argued that state intervention would only stifle innovation.

James Carnes, co-founder of emissions-tracking firm Iconic Air, said companies should be left to self-police, arguing that federal regulations would hamper growth and that fears of rogue AI models are driven largely by misinformation. Andrew Wyatt, co-founder of the AI-powered media monitoring firm Delve, compared government regulators to airport security agents who slow down processes that businesses could manage more efficiently themselves.

Yet some attendees agreed with the governor's caution. Eser Saydam, a senior system engineering manager at Nutanix, said the current pace of AI development is "reckless," while Ameena Elahi, assistant director of AI initiatives at Penn Medicine, warned against rushing ahead "at the expense of people."

The political battleground

The debate carries sharp political undertones. Shapiro, widely viewed as a 2028 Democratic presidential contender, used his speech to criticise House Speaker Mike Johnson for adjourning the House early, and Donald Trump, who recently dismissed apocalyptic warnings about AI as a "hoax".

In response, Trump administration spokesperson Liz Huston said the White House is committed to "unleashing American innovation" and supporting the deployment of advanced AI models without heavy-handed guardrails.

Shapiro also framed the issue as a geopolitical race, stating, "We need to beat China on AI." However, he suggested building an "international consensus" on safety, pointing to an upcoming US visit by Chinese President Xi Jinping as an opportunity to start talks.

Back home, Shapiro’s shifting stance has drawn attacks. His Republican opponent in the upcoming November election, Stacy Garrity, called his comments a "shameless flip-flop," pointing to his past support for the AI and data centre industries. Recent polls show Shapiro holding a 16-point lead over Garrity.

Local ambitions and deficits

Earlier in the summit, Pittsburgh Mayor Corey O'Connor pitched the city as an AI hub to help plug a projected municipal budget deficit of more than $30 million.

But reflecting local unease about the technology's footprint, O'Connor ruled out hosting heavy data centres within city limits or using AI to replace municipal workers. Instead, the mayor suggested the technology should stick to minor tasks, such as tracking potholes, streamlining permits, and coordinating winter snowploughs.