Anthropic Writes Its Own AI Safety Bill — Government Could Block Models Only in Catastrophes

Anthropic, the company behind the Claude AI assistant, has quietly written its own draft AI safety bill and is pitching it to members of the US Congress, Politico reported on Friday. The unusual move — a tech company drafting the law it wants to be governed by — has set off a fierce debate about who should write the rules for artificial intelligence.

The draft sets an extremely high bar for government intervention. A powerful AI model could be blocked from release or use only if it posed an imminent risk of causing more than $1 billion in damages, or of killing or seriously injuring more than 50 people. And the danger would have to come from just three sources: the creation of a biological weapon, a cyberattack, or a model escaping its developer’s control.

Under the proposal, independent evaluators registered with the Commerce Department would regularly inspect frontier labs and their most powerful models, reporting any imminent risk within 72 hours. Developers would also have to test their systems, publish the findings and maintain strong security programs.

An Anthropic spokesperson confirmed the draft reflects the company’s “Advanced AI Framework” safety blueprint. Critics argue the thresholds are so high that government intervention would almost never happen.

Source: Full story

Comments

Popular posts from this blog

Fairbairn's Walk-Off Field Goal Seals Wembley Thriller as Texans Edge Jaguars 34-31

Anthropic IPO: Public S-1 Filing Expected Today Ahead of $2 Trillion Listing

Netflix Beats Earnings but Slashes Ad Forecast — Stock Tumbles After Hours