Trump’s AI Pact: What to Know | National News

Americans’ anxiety about artificial intelligence is real – and it’s rising. But when President Donald Trump met this week with top tech executives, he insisted that the growth of AI will be a boon to the U.S. economy and that the companies can police themselves.

“We do not surrender to fears about what technologies will do to us,” said Trump, who has dismissed worries about AI as a “hoax.”

Taking a largely hands-off approach, Trump signed a pact with the executives in which they voluntarily agreed to develop “robust internal controls” to stop AI agents from going awry. But the pact leaves many of the public’s concerns unresolved.

Trump’s enthusiasm for AI, despite warnings from some industry leaders, raises questions about what would have to happen for the government to take a more active role. Here’s a look at the public’s concerns and what the agreement might mean.

In a clear sign of public discontent with the rapid expansion of AI, voters in communities around the country have been pushing back against efforts to build massive data centers needed to run the systems. But polls show that workers and consumers have much wider fears.

More than half of Americans say they are more concerned than excited about AI, up from 37 percent in 2021, according to the Pew Research Center. Those worries are particularly pronounced among young adults just starting their working lives. About three of every four young adults say that AI will lead to fewer jobs over the next two decades, a notable increase.

In another recent poll, this one by Politico, nearly two-thirds of Americans expressed moderate to grave concerns that AI could someday destroy humanity. And about half of all Americans say data centers are bad for the environment, home energy costs and the quality of life, all up significantly over the last two years, according to Pew.

At just over 300 words, the agreement acknowledges fears about AI but avoids imposing government guardrails, declaring that “every company is responsible for developing its own technology safely.”

That is notable given recent warnings by some tech leaders that the pace of AI development should be slowed to protect the public from risk.

The agreement leaves those safeguards largely up to the companies “training and deploying frontier (AI) technologies.” But the executives who signed the pact pledged to put measures in place to make sure the technology performs as intended. Those include developing internal controls to monitor AI models and their impact on cybersecurity, biosecurity and chemical threats. The companies also agreed to staff internal teams to keep watch and partner with auditors to independently assess their progress.

The agreement was signed by top executives of six U.S. companies leading the development of AI: Google’s Sundar Pichai; Dario Amodei of Anthropic; Mark Zuckerberg of Meta; OpenAI President Greg Brockman; Elon Musk of xAI; and Nvidia’s Jensen Huang. Trump also signed. When he posted a copy of the signature page to social media afterward, it included a spelling error, listing him as “President of the Unites States.”

Despite calls by some members of Congress, the agreement rolled out by Trump steers clear of government controls on AI development. Instead, it says only that “over time, it may make sense to codify these steps into laws or regulations.”

Trump said all the parties believed in self-regulation and that the agreement was a form of protection. Asked whether there shouldn’t be formal rules governing AI, he said industry players “love our country and the world,” and that was enough. While it’s up to the companies to take action, Trump said the agreement is “morally binding.”

With the White House playing up the benefits of AI, Trump announced he was moving to rebrand the technology. From now on, the federal government will refer to it as “super intelligence” or SI, according to an executive order Trump signed on Tuesday.

That term, the order says, better reflects the promise and potential of the technology. The rebranding effort notwithstanding, Democrats in Congress and many voters are skeptical of Trump’s approach to AI.

Just a quarter of those surveyed recently by Quinnipiac University approved of his handling of AI, with 58 percent disapproving.