Bill would require annual AI audits and appeal rights

What does this do?

This bill would require any company or government agency using AI to make important decisions about hiring, housing, loans, health care, education, criminal justice, or voting to test those systems for bias before use and every year after, using an independent reviewer. Companies would have to post plain-language explanations of how their AI works, what data it uses, and what audits found. It would also be illegal to punish anyone who raises concerns about an AI system.

Who does it affect?

Any company or government agency using AI or complex computer systems for major decisions about people would have to follow these rules. People affected by AI-driven decisions would gain the right to request a human decision-maker and to appeal to a human reviewer.

Why does it matter?

Without these requirements, AI systems used in high-stakes decisions could go unchecked for bias or errors that affect people's lives. The Federal Trade Commission would enforce the rules, and individuals and state governments could also take legal action against violators.

Where does it stand?