President Donald Trump’s approach to artificial intelligence regulation appears to be shifting as concerns about AI safety, cybersecurity, and competition with China continue to grow. According to recent reporting from NPR, the Trump administration initially focused heavily on accelerating AI innovation and removing regulatory barriers. Early actions included revoking parts of the Biden administration’s AI safeguards and promoting a more industry-friendly approach designed to help American companies compete globally.
In recent weeks, however, the White House has signaled a possible change in tone. Trump recently stated that “there should be regulations on AI,” reflecting growing concern about the risks posed by increasingly advanced AI systems. Much of this discussion intensified after AI company Anthropic warned that its powerful new AI model, Mythos, could create serious cybersecurity threats if misused. Officials inside the administration are now debating whether AI testing and oversight should become more formalized or remain largely voluntary.
The debate has exposed divisions within the administration itself. Some national security officials reportedly want intelligence agencies to play a stronger role in evaluating advanced AI models because of fears about cyberattacks, infrastructure threats, and foreign competition. Others, especially within the Commerce Department and parts of Silicon Valley, continue to support a lighter regulatory framework that prioritizes innovation and economic growth. The administration is also pushing for a national AI framework that could limit individual state regulations and establish a more unified federal approach.
Despite the tougher rhetoric, Trump’s overall AI strategy still appears more business-friendly than many international regulatory models, especially compared with Europe’s stricter rules. The administration continues to emphasize maintaining American leadership in artificial intelligence while avoiding regulations that could slow development. Still, the growing conversation around AI safety, cybersecurity, and national security suggests that even strong supporters of rapid AI expansion are beginning to recognize the need for at least some level of oversight as the technology becomes more powerful.
