Anthropic AI bill lets US block models that cause over ₹9,496 cr in damage

Anthropic pitched US lawmakers on a draft bill to regulate frontier AI safety and security, Politico reported. The bill would allow government to block AI model's release or use only when it poses imminent, foreseeable risks, including over $1 billion (₹9,496 crore) in damages or serious harm to over 50 people. It also mandates safety testing, transparency and independent evaluations.