What Anthropic's Leaked AI Bill Means for Players

If passed, this bill would set a high bar for government intervention, potentially limiting oversight to only the most catastrophic AI risks while leaving everyday harms unaddressed.

If passed, this bill would set a high bar for government intervention, potentially limiting oversight to only the most catastrophic AI risks while leaving everyday harms unaddressed.

Key points - Anthropic's leaked draft bill would trigger government intervention only at $1 billion in damages or 50+ injuries/deaths. - Risks must stem from biological weapons, model evading developer controls, or cyberattack. - Anthropic says the bill reflects its Advanced AI Framework, advocated since June. - Critics argue such regulations give tech companies appearance of capitulation while restricting on their own terms.

Timeline - June 2024: Anthropic begins advocating its Advanced AI Framework to federal and state lawmakers. - Earlier this year: Anthropic's safety lead quits, stating 'the world is in peril'. - Last week: Eight US states' attorneys general accuse OpenAI of ignoring safety warnings for profit.

FAQ

Q: What does the leaked Anthropic bill actually propose? A: It would allow government intervention only if an AI model poses an imminent risk of over $1 billion in damages or killing/seriously injuring more than 50 people from biological weapons, evading controls, or cyberattacks.

Q: Why are people skeptical of this bill? A: Critics argue it lets AI companies self-regulate on their own terms, creating an appearance of safety without addressing existing harms.

Q: Has this bill been introduced in Congress? A: No, it's a confidential draft that Anthropic has pitched to lawmakers, but it has not been formally introduced.