On March 13 the United States Senate voted 56‑43 to approve a $500 million appropriation for the newly proposed National AI Safety Institute. The institute will conduct basic and applied research into robust, transparent and trustworthy AI systems, develop best‑practice guidelines, and provide independent oversight of emerging technologies.
Senator Maria Gonzales (D‑TX), the bill’s primary sponsor, said the funding would “ensure that the nation remains at the forefront of safe AI while protecting citizens from unintended harms.” The measure also directs a portion of the money to an international consortium that will share safety protocols with allies.
"We cannot afford to be left behind on safety as we race toward the next generation of AI," Senator Gonzales declared during the floor debate.
The legislation follows a series of executive‑order‑driven initiatives, but it marks the first time Congress has committed substantial public funds specifically for AI risk mitigation. Experts predict that the institute could accelerate the deployment of safety‑enforced models in critical sectors such as healthcare, finance and national defense.