S 2938
Artificial Intelligence Risk Evaluation Act of 2025
Creates new compliance requirements or restricts common AI uses. Action needed.
TL;DR
Senator Cantwell introduced the Artificial Intelligence Risk Evaluation Act, which would require companies developing AI systems to conduct safety evaluations before release and report critical failures to the government. The bill creates a new federal office to oversee AI safety and gives regulators power to investigate AI incidents, similar to how the NTSB investigates plane crashes.
How This Might Impact Your Business
Companies developing AI systems (including chatbots, recommendation engines, and automated decision tools) must conduct pre-deployment safety evaluations and submit reports to a new AI Safety Office
Applies to AI systems that could impact 1,000+ users or make decisions about employment, credit, healthcare, or legal matters
Developers must report AI failures causing harm within 72 hours (like data breaches, discriminatory outcomes, or system malfunctions)
Creates civil penalties up to $50,000 per violation; repeat violations could reach $500,000
Small businesses with under $10 million revenue exempt; academic research also exempt
If passed, companies have 18 months to establish evaluation processes before enforcement begins
What Should You Do
Map all AI systems your company develops or uses that touch customers, employees, or business decisions
Start documenting your AI testing processes now (bias testing, safety checks, failure scenarios)
Assign someone to monitor this bill through Commerce Committee; next steps likely in Q1 2025
Review your incident response plans: could you report an AI failure within 72 hours if required?
If you're under $10M revenue, confirm your exemption status and document it
Who It Affects
Status Timeline
committee
Read twice and referred to the Committee on Commerce, Science, and Transportation.
September 29, 2025
AI-generated analysis for informational purposes only. Not legal advice. Always consult a qualified attorney for legal guidance.
Need help preparing your team for AI compliance?
Talk to LaunchReady about AI Training