Key facts
- US Senate negotiators are debating legislation requiring AI companies to demonstrate reasonable precautions against potential harm.
- The proposed legislation includes a "duty of care" requirement for AI developers.
- The U.S. commerce secretary could be empowered to request evidence of these precautions and deploy auditors.
- Senator Amy Klobuchar is working on a bipartisan agreement for government oversight of AI risks.
- Klobuchar stated the agreement would require developers to work with government experts to verify and test AI models for safety.
- U.S. President Donald Trump believes existing authorities are adequate for regulating tech companies.
U.S. Senate negotiators are considering legislation that would mandate AI companies to prove they are taking adequate steps to prevent their technologies from causing harm, according to sources familiar with the discussions. The proposed measures, which could grant the U.S. commerce secretary the authority to demand evidence of a "duty of care" and to deploy government auditors for product testing, are still under debate.
Concerns about AI's potential risks have been amplified recently, particularly after an Anthropic researcher resigned, citing beliefs that AI could pose an existential threat by the end of the decade. This sentiment was echoed by leaders of major AI companies who called for a slowdown in the technology's development, leading to a dip in AI-related stocks globally.
The legislative effort is being spearheaded by Senate Majority Leader John Thune, Senate Commerce Committee Chairman Ted Cruz, and Senator Amy Klobuchar, who is leading the Democratic involvement. Klobuchar confirmed her commitment to achieving a bipartisan agreement on AI oversight, emphasizing the need for developers to collaborate with government experts to ensure AI safety through verification and testing.
However, the path for this legislation to become law appears challenging. U.S. President Donald Trump has suggested that current legal frameworks are sufficient for regulating tech companies, indicating a potential reluctance to sign new AI-specific rules.