Story perspectives

Anthropic Draft Bill Would Let Government Block Dangerous AI Models

By Drooid · Oct 18, 2026, 10:02 PM UTC · How we work

Summary

Story summary

Anthropic PBC presented a draft bill to Congress to regulate AI model releases.
The bill would allow the government to block a model causing over $1 billion in damages.
Blocking requires a foreseeable, imminent risk of killing or seriously injuring more than 50 people.
The Commerce Department can order a model’s suspension.
Anthropic spent more on federal lobbying in the first half of 2026 than in all of 2025.

History

Historical Context

The EU adopted the GDPR on 14 April 2016, effective 25 May 2018, and has levied fines of up to €20 million (4 % turnover) on firms such as Google and Facebook. Seen as the “Brussels effect,” the regime now parallels Anthropic’s 2026 draft bill allowing the U.S. Commerce Department to suspend AI models that could cause $1 billion damage, marking a new trans-Atlantic oversight push.

1 of 2

Story summary

Anthropic PBC presented a draft bill to Congress to regulate AI model releases.
The bill would allow the government to block a model causing over $1 billion in damages.
Blocking requires a foreseeable, imminent risk of killing or seriously injuring more than 50 people.
The Commerce Department can order a model’s suspension.
Anthropic spent more on federal lobbying in the first half of 2026 than in all of 2025.

1 / 2