Anthropic, the company behind Claude models, has submitted a confidential draft bill on frontier AI safety to US lawmakers. The proposal would give the federal government the power to halt an AI model’s deployment when potential damages exceed $1 billion or when the AI could cause death or serious injuries to more than 50 people. The draft targets three specific danger zones: biological weapons, a developer losing control of its own model, and cyberattacks. Independent evaluators would need to register with the Commerce Department, and emergent risks would have to be reported within 72 hours. The proposal mirrors California’s SB 53 law from 2025 but remains confidential and has not been formally introduced.
Source: Read the original article
Disclaimer: this content is for information purposes only and is not financial advice. Cryptocurrencies are highly volatile: you may lose all of your capital. Always do your own research. Legal notice
