U.S. President Donald Trump confirmed on Sunday, September 27, 2026, that he planned to hold a dinner meeting with Anthropic CEO Dario Amodei. Despite the meeting, Trump reaffirmed his opposition to slowing the development of artificial intelligence, stating that maintaining a technological lead over China is a primary national security interest.
The meeting followed a joint call on Saturday, September 26, 2026, by the CEOs of Anthropic, OpenAI, and xAI to "pace the frontier" by slowing the improvement of model capabilities. Amodei, Sam Altman of OpenAI, and Elon Musk of xAI advocated for a slowdown to allow for the implementation of safety measures. Their statements came shortly after Anthropic reported that its Claude models had been targeted or misused for activities including attempted bioweapons development and cyber operations.
Trump, speaking from the Presidents Cup golf tournament in Illinois, dismissed concerns about "rogue" AI agents and described recent warnings as a "hoax." He stated that the U.S. currently holds a lead of approximately 18 months over China in AI development. In a separate interview on NBC’s "Meet the Press," Microsoft co-founder Bill Gates countered Trump's characterization, arguing that AI is powerful enough to cause "a billion deaths" and that federal safeguards and monitoring are necessary.
House Speaker Mike Johnson (R-LA) stated that while guardrails are needed, Congress should not "rush" into regulation that could impact global competition. Conversely, Secretary of Transportation Pete Buttigieg argued that Congressional inaction is dangerous given the speed of technological advancement. Meanwhile, David Sacks, co-chair of the President's Council of Advisors on Science and Technology, suggested the leading companies should slow their own development voluntarily rather than seeking government intervention that might impact smaller competitors.
For ordinary citizens, the outcome of this policy debate could determine the security of their digital interactions and the speed at which AI-driven changes affect their jobs or health care. If Amodei’s warnings are accurate, he suggests "swarms of rogue AI agents" could impact the internet within six months without intervention. Conversely, Trump’s position suggests that a slowdown would result in a loss of American technological dominance, potentially allowing foreign competitors to set global standards for the technology.
The precedent set by this interaction highlights a shift where industry leaders are actively requesting government oversight, a move David Sacks characterized as a potential "cartel" behavior intended to freeze out open-source developers. What happens next depends on whether Congress acts before the November elections. House Speaker Mike Johnson indicated he does not plan to cancel recess for AI legislation but proposed a future meeting with platform providers to discuss potential guardrails. Anthropic and OpenAI have both pledged to allow external reviewers employee-level access to their systems to monitor for misuse.