Amazon-backed (AMZN) Anthropic shared several policy proposals Wednesday, calling on the US Congress to not preempt state law without enacting a federal law to address risks related to artificial intelligence.
Anthropic said that while several recent state laws require companies to describe their safety practices and share them publicly, the rapid pace of AI acceleration warrants a more substantial role from governments.
Anthropic highlighted several key requirements as part of its proposed framework for frontier developers, including transparency, qualified independent evaluation, security against cyberattackers, and the deployment of a regulatory authority that can block or deter models that pose significant risks.
The rules are intended to address biological risks, cyber risks, risks related to loss of control, and automated research and development risks, Anthropic said.
Comments