"AI systems should only be shut down if they cause more than $1 billion in damage or result in more than 50 people getting hurt or killed," said Anthropic in a proposal to the U.S. Congress.
Anthropic has proposed a draft law to the US Congress that would allow the government to block the release or use of AI models only in cases where there is an imminent risk of over $1 billion in property damage or more than 50 casualties. The bill, which aims to regulate frontier AI, outlines specific scenarios where intervention is warranted, including the creation of biological weapons, uncontrolled AI behavior, and cyber attacks.
The US Department of Commerce would be responsible for evaluating AI models and could request a court to halt the release of a model deemed risky. The proposed law would also limit the ability of other federal agencies to halt AI development, distribution, or internal use.
Written by urgent.news from Hankyoreh's report — not a translation of it. Machine-written — may contain errors; check the original before relying on it.
