Regulation
US Senate advances AI liability bill targeting frontier labs
The bipartisan SAFE AI Act would make frontier model developers legally liable for harms caused by their models in high-risk domains.
The Senate Commerce Committee voted 14-11 to advance the SAFE AI Act, which would impose strict liability on AI developers whose models cause harm in healthcare, financial services, and critical infrastructure. The bill includes a safe harbor for companies that complete third-party audits. Industry groups have pushed back, arguing the liability standard would chill open-source development.
Source
Reuters · reuters.com