Federal AI Preemption, Anthropic’s Enterprise Surge, and OpenAI’s $840B Bt: AI Policy Week in Review

Three developments this week signal a pivotal moment for AI in the United States: the White House is moving to consolidate AI oversight at the federal level, Anthropic has staged a remarkable enterprise market takeover, and OpenAI has closed one of the largest private funding rounds in tech history. Taken together, they sketch an industry racing forward on multiple fronts simultaneously.

White House Pushes Federal Preemption of State AI Laws

On March 20, 2026, the Trump administration released a comprehensive AI policy framework urging Congress to preempt state-level AI regulations, according to reports from the LA Times and the National Law Review. The framework directly targets legislation from states including California, Colorado, Utah, and Texas, arguing that a patchwork of state rules risks stifling innovation and undermining U.S. global AI leadership.

The administration’s position contrasts sharply with the European Union’s rights-centered AI Act approach, emphasizing competitiveness over consumer protection as the primary policy lens. The implications are significant:

• For AI developers: A single federal standard could reduce compliance overhead across state lines.
• For state regulators: Years of legislative work in states like California could be superseded.
• For civil society: Advocates warn a weaker federal floor may leave consumers with fewer protections.

The debate over federal AI preemption is likely to define the U.S. regulatory landscape for the next several years, and Congress’s response will be closely watched by both industry and international partners.

Anthropic Captures 73% of New Enterprise AI Buyers

In perhaps the most striking market data point of the week, Anthropic now accounts for over 73% of new corporate AI buyers, according to spending data from Ramp. That figure is remarkable given that just 10 weeks earlier, Anthropic and OpenAI were splitting the enterprise market roughly 50/50.

The shift demonstrates how quickly enterprise AI adoption patterns can move. Anthropic’s Claude models have gained traction for their strong performance on long-context and