The United States is dangerously late in establishing regulatory guardrails for artificial intelligence, according to prominent investors and technology experts, even as the industry itself increasingly acknowledges the need for government oversight to protect public safety and security.
Paul Tudor Jones, speaking on CNBC's "Squawk Box" on Thursday, warned that regulatory action should have already occurred. "We need to do it tomorrow. We're late already. We should have already done it," Jones said, underscoring the urgency with which experts view the regulatory gap. His comments reflect a striking shift in industry sentiment: at a recent conference with AI experts and model makers, 80% of participants supported AI regulation, up from around 20% last year. One company leader at the conference expressed surprise that the industry was not yet regulated.
The absence of comprehensive federal AI regulation in the United States stands in sharp contrast to international action. The European Union passed the AI Act two years ago, establishing a framework for managing risks associated with artificial intelligence deployment. Meanwhile, some U.S. states have passed or introduced their own legislation, many of which have targeted child safety—a reactive approach that leaves significant gaps in protection across sectors and jurisdictions.
The Regulatory Lag and Public Protection
Lawmakers and experts have long advocated for regulations to mitigate safety, privacy and security concerns associated with artificial intelligence technology. These concerns span multiple domains: the potential for AI systems to amplify discrimination, the risks posed by deepfakes and synthetic media, data privacy vulnerabilities, and the concentration of power in the hands of a small number of companies controlling the most advanced models.
In March, the White House released a nationwide AI policy framework, signaling federal recognition of the issue. However, the framework represents a starting point rather than comprehensive regulation. The gap between the framework's announcement and the absence of binding regulatory requirements leaves significant uncertainty about how—and whether—federal standards will be enforced.
The patchwork of state-level legislation, while addressing specific harms like child exploitation, fails to establish consistent national standards. This fragmentation creates compliance challenges for companies and leaves many areas of AI deployment entirely unregulated, allowing market forces rather than democratic oversight to determine how the technology develops and is deployed.
International Competition and Safety Dialogue
Jones noted that the United States is engaged in a heated rivalry with China to produce the most advanced AI models and strategy. However, he argued that this competition should not preclude cooperation on safety standards. "Everyone wants what's best for their people," Jones said, adding that he does not believe China wants to "wipe out" the U.S. Instead, he advocated for direct engagement: "We should be having a dialogue with them about AI safety."
This perspective reflects growing recognition among experts that AI safety transcends national borders and competitive interests. The Wall Street Journal reported this week that both countries are considering official discussions about AI at an upcoming meeting between President Donald Trump and China's Xi Jinping, suggesting that high-level diplomatic engagement on the issue may be imminent.
The framing of AI regulation as both a domestic imperative and an international cooperation challenge highlights a critical tension: the technology's development is driven by competitive dynamics between nations, yet its risks—from deepfakes to privacy breaches to systemic bias—affect populations globally. Jones's proposal for watermarking AI-generated content to distinguish it from authentic material exemplifies the kind of technical and policy solutions that require coordinated action across jurisdictions.
Why This Matters:
The regulatory lag in the United States reflects a pattern where market-driven technological development has outpaced democratic governance. With 80% of AI industry participants now supporting regulation—a dramatic shift from 20% support last year—the consensus exists for action, yet federal policy remains fragmented and incomplete. The absence of comprehensive regulation means decisions about how AI systems are trained, deployed, and audited are made primarily by private companies rather than through democratic processes accountable to the public. This gap has concrete consequences: workers lack protections against algorithmic discrimination in hiring, consumers face risks from deepfakes and manipulated content, and marginalized communities bear disproportionate risks from biased AI systems. The fact that the European Union established binding regulations two years ago while the U.S. remains in a framework stage underscores how regulatory delay allows harms to accumulate. International cooperation on AI safety, as Jones advocates, is essential—but it cannot substitute for domestic regulatory authority that establishes enforceable standards protecting American workers, consumers, and democratic institutions.