Anthropic's AI Risk Bill Proposes $1 Billion Damage Threshold
Anthropic has drafted a bill proposing a $1 billion damage threshold to define severe artificial intelligence risks. This proposal has sparked debates among policymakers regarding the impact of federal regulations on state laws and the adequacy of the regulation.
AI company Anthropic has submitted a draft legislative proposal that includes a $1 billion damage threshold to identify severe artificial intelligence risks. While the company argues that this threshold is a tool to separate public and national security risks from short-term issues, industry experts and policymakers express differing views on the effectiveness of this limit and its effects on federal preemption.
The $1 billion damage threshold included in the draft text is evaluated by some experts as a high bar, though it is noted that AI-induced damages could rapidly reach this level. Chamber of Progress CEO Adam Kovacevich suggests that Anthropic is using this definition to ensure federal regulations supersede state laws.
The Democratic wing worries that a federal standard could weaken states' robust AI laws. Anthropic representatives defend the draft as the strongest option, stating that they will continue to work on the text based on feedback. The current proposal is considered weaker by some circles when compared to other legislative initiatives like the FRONTIER Act.
Sources
- politico.com · Link