Despite growing concerns from industry leaders and experts about the risks posed by artificial intelligence (AI), efforts to establish comprehensive federal regulations in the United States have largely stalled amid political and technical challenges. Over the past two years, lawmakers have proposed numerous bills aimed at addressing safety, national security, and economic issues related to AI, but partisan divisions and industry opposition have hindered significant legislative progress.
Several proposed measures seek to institute federal oversight of advanced AI systems. These include requirements for companies to report safety incidents, undergo independent audits, and potentially subject their technologies to government suspension if they pose catastrophic risks. The Frontier Act, introduced by Representative Jay Obernolts, a Republican from California, and Representative Lori Trahan, a Democrat from Massachusetts, is among the most notable, though it has not advanced in Congress. It also includes provisions to pre-empt state-level regulations, a point of contention among states eager to act amid federal delays.
On the Senate side, Democrat Amy Klobuchar of Minnesota and Republican John Thune of South Dakota are working on a bill expected to focus on extreme risks involving biological or nuclear threats. Despite some cooperation across party lines, few bills have moved beyond introduction or committee stages.
One widely discussed idea gaining bipartisan interest is the implementation of a "kill switch" to allow the government to deactivate AI models that threaten public safety. While the technology’s feasibility remains debated, figures like Senator John Kennedy, Republican of Louisiana, and Representative Anna Paulina Luna, Republican of Florida, have advocated for it. Luna has cited a lack of technical expertise among older lawmakers and concerns about competition with China as factors contributing to the slow pace of regulation.
National security concerns also loom large. Legislative proposals aim to safeguard critical infrastructure and limit adversaries’ access to sensitive AI technology, including chips essential for AI development. For instance, the Advanced AI Security Readiness Act would task the National Security Agency with devising security plans against cyber intrusions. Additional bills seek to restrict AI-enabled weapons systems by mandating human oversight before their deployment.
The economic ramifications of AI-driven automation have prompted lawmakers to consider protections for American workers and measures to manage rising utility costs associated with data center power demands. Proposals range from using AI companies’ profits to fund retraining programs to incentivizing energy regulators to prevent rate hikes. However, these initiatives have yet to secure enough support for enactment.
On the executive level, former President Donald Trump has expressed opposition to stringent AI regulation, describing concerns about AI’s harms as a "hoax," and instead emphasizes accelerating AI innovation to bolster U.S. economic and technological leadership. Under his administration’s later directives, government reviews of AI models remain voluntary. However, the Department of Commerce has shown willingness to employ existing export control laws to block certain AI technologies deemed risky, such as temporarily banning Anthropic’s Mythos 5 and Fable 5 models in mid-2026 due to cybersecurity concerns.
Congressional leaders have acknowledged the importance of addressing AI but have struggled to find consensus and effective enforcement mechanisms. Speaker Mike Johnson, aligned with Trump, advocates for corporate responsibility rather than legal mandates, whereas House Minority Leader Hakeem Jeffries, a Democrat, criticizes Republican reluctance to act.
Internationally, China has taken a different approach, rejecting calls for a pause in AI development despite updating its AI safety governance framework. Chinese companies like DeepSeek and ZAI continue to advance AI capabilities, narrowing the technological gap with U.S. firms. Upcoming discussions between President Trump and Chinese President Xi Jinping are expected to touch on AI, though meaningful cooperation appears limited given their mutual interests in technology dominance.
With midterm elections approaching and public concern over AI safety and economic impacts rising, lawmakers face increasing pressure to move forward. Nevertheless, experts anticipate that significant regulation is unlikely until a major disruptive event compels swift action, suggesting that a broad safety pause in AI development remains improbable for now.