A bipartisan AI safety framework in the US has been postponed until after the Senate recess. The bill, backed by Senators John Thune, Ted Cruz, and Amy Klobuchar, would require AI developers to notify the Commerce Department of significant AI capability breakthroughs and grant the government authority to seek court injunctions against non-compliant companies. Anthropic uniquely supports the bill while simultaneously calling for stricter requirements, particularly mandatory public disclosure of AI risks. Meanwhile, Senator Maria Cantwell opposes the bill’s approach and favors relying on expertise from the Commerce Department and national labs to evaluate AI technologies.
Source: Read the original article
