The US federal AI landscape split this week. The White House released its national AI policy framework on March 20, calling on Congress to preempt state AI laws it views as burdensome to innovation. Within days, Sen. Marsha Blackburn (R-TN) introduced a competing measure, the TRUMP AMERICA AI Act, that reportedly takes the opposite approach.
The two frameworks disagree on the questions that matter most to compliance teams right now.
On state law authority: the White House framework explicitly pushes for federal preemption, which would clear the current patchwork of state AI laws. The Blackburn bill, according to reporting, does not include preemption, meaning state laws would remain in force alongside any federal requirements.
On copyright: the White House framework reportedly leaves AI training and fair use questions to the courts. The Blackburn bill reportedly declares AI training on copyrighted works categorically outside fair use, a definitively harder line. The IAPP confirmed coverage of the bill’s focus on copyright and child protections, though specific provision details remain sourced to outlets that could not be fully verified at publication time.
One note on the Blackburn bill’s status: it’s a discussion draft, not yet formal legislation. Reported as roughly 291 pages, it also reportedly includes a federal products liability framework for AI systems, mandatory annual third-party bias audits for high-risk AI, and age-verification requirements for minors, though those specific provisions require qualification pending source confirmation.
For compliance teams, the practical takeaway is uncomfortable: there’s no single federal framework to align to. Both visions are active, both have congressional backing, and they point in different directions. Planning for state law persistence while monitoring federal preemption risk is the only defensible posture right now.