Overview
- OSTP Director Michael Kratsios testified that the 2026 AI Action Plan aims to pair rapid innovation with foundational safety standards, but he declined to outline what a preemptive national framework would include or when recommendations will arrive.
- He cited the December executive order that directs agencies to challenge state AI laws deemed onerous and to condition certain federal funds, and that tasks him and David Sacks with drafting legislative recommendations.
- Kratsios emphasized voluntary, sector-specific guardrails through NIST and its Center for AI Standards and Innovation, as the administration presses revisions to NIST’s AI Risk Management Framework to remove references to misinformation, DEI, and climate change.
- Lawmakers flagged operational hurdles, including a GAO-cited 40% shortfall in cleared AI talent; Kratsios acknowledged the gap and described efforts to streamline hiring and expand rotations between industry and government.
- Debate over federal–state authority continued, with Rep. Zoe Lofgren calling the executive order unconstitutional, Chairman Jay Obernolte urging clear federal and state roles, and members raising concerns about data centers’ strain on electric grids as the White House pressures companies to avoid higher local bills.