The race to regulate AI has sparked a federal vs. state showdown (techcrunch.com)

A high-stakes federal vs. state showdown is unfolding over who gets to regulate AI. In the absence of a unified federal consumer-safety standard, dozens of states have passed more than 100 AI-related laws this year (38 states active), targeting deepfakes, transparency and government AI use — though a study found 69% impose no developer requirements. Tech firms and well-funded pro-industry PACs (e.g., Leading the Future has raised $100M+) argue state-by-state rules create an “unworkable patchwork” that will stifle innovation and slow U.S. competitiveness with China. In response, Congress and the White House are exploring preemption: House negotiators have attempted to tuck anti-preemption language into the NDAA, and a leaked draft executive order would create an “AI Litigation Task Force” to challenge state laws and steer agencies toward national standards — with entrepreneur David Sacks positioned to co-lead the effort. The clash matters because it will shape whether regulation emphasizes rapid innovation or proactive safety and accountability. Rep. Ted Lieu and the bipartisan House AI Task Force are drafting a 200+-page federal “megabill” covering fraud, deepfakes, child safety, whistleblower protections, compute access for academics, and mandatory testing/disclosure for large language models — a lighter touch than rival proposals calling for government-run predeployment evaluations. More than 200 lawmakers and nearly 40 state attorneys general have pushed back against broad federal preemption, arguing states are “laboratories” for policy. Expect protracted legal and legislative battles that will determine if U.S. AI governance is centralized, industry-led, or a fragmented mosaic of state rules — with major implications for labs, startups, consumers, and liability regimes.