WASHINGTON — The Senate Commerce Committee voted 18-7 on Thursday to advance the Artificial Intelligence Accountability Act, a landmark bipartisan bill that would create the first comprehensive federal regulatory framework for AI systems deployed in high-stakes applications.
The legislation, co-sponsored by Senators Maria Cantwell (D-WA) and John Thune (R-SD), would establish mandatory safety testing requirements for AI systems used in healthcare, financial services, employment decisions, and critical infrastructure. It would also create a new office within the Commerce Department tasked with monitoring AI development and enforcing compliance.
"This is not about stifling innovation — it's about ensuring that as AI becomes more powerful, we have guardrails in place to protect American consumers and workers," said Cantwell at a press conference following the vote. "The goal is responsible development, not development at any cost."
The bill's advancement comes amid growing concerns about the rapid deployment of AI systems in sensitive domains, including documented cases of algorithmic bias in hiring platforms and errors in AI-assisted medical diagnostics. A recent study by the National Institute of Standards and Technology found that approximately 34% of AI systems used in consumer-facing applications exhibited some form of discriminatory behavior.
Notably, the legislation has garnered support from major technology companies, marking a shift from the industry's traditional resistance to regulatory oversight. Representatives from Microsoft, Google, and Anthropic testified in favor of the bill during committee hearings, arguing that clear federal standards would provide regulatory certainty and help prevent a patchwork of conflicting state laws.
"The question is not whether AI will be regulated, but how," said Brad Smith, Microsoft's Vice Chair, in prepared testimony. "Industry prefers to engage constructively with thoughtful federal legislation rather than navigate 50 different state regimes."
Consumer advocacy groups have also endorsed the framework, though some argue it doesn't go far enough. The Center for AI Safety called for stronger enforcement mechanisms and mandatory algorithmic audits by independent third parties, provisions that were diluted during committee negotiations.
The bill now moves to the full Senate for consideration, where supporters are confident it will pass with bipartisan support. If approved, it would then need to clear the House, where a companion bill has already been introduced with 47 co-sponsors.