Draft AI safety bill from Anthropic sparks backlash

Rival OpenAI warns no single company should write rules for such technology

By LineZotpaper

Published20 October 2026

Read Time2 min

A leaked draft of an AI safety bill proposed by Anthropic to lawmakers on Capitol Hill has triggered backlash from safety advocates and a pointed response from rival OpenAI, with critics arguing the bill does not go far enough and unfairly pre-empts state laws.

The debate over federal AI safety regulation intensified after a draft proposal from Anthropic, long viewed as a safety-focused frontier lab, was leaked late last week. The bill, which a company spokesperson said is a legislative version of an AI safety framework Anthropic has been pitching for months, has put the company on the defensive. One of its top lobbyists posted about planned tweaks to better reflect the pace of AI progress and the company's stance on pre-empting state laws.

AI safety advocates have pushed back, arguing that the draft does not go nearly far enough, unfairly pre-empts state laws, and accepts too much potential risk.

OpenAI quickly sought to distinguish itself from its rival. Chief Global Affairs Officer Chris Lehane wrote in a post Monday implicitly directed at Anthropic: "Policy decisions must ultimately be made through our democratic processes with our elected officials leading. … No one company (including OpenAI) should write the rules alone for a technology with such profound implications for our economy, society and national security."

Anthropic has increasingly embraced regulation in a way other AI companies have not. As the risks of its advanced AI have become apparent and its tensions with the US government have grown, the company has pursued regulation through methods that have sparked debate within the federal government, including pitching terms that would allow officials to block the release of certain models. OpenAI has generally been a more cautious Washington player, seeking to endorse compromise bills written by lawmakers while encouraging similar frameworks across the states.

Analysis

Why This Matters

The leaked bill reveals a key player's regulatory strategy, potentially shaping federal AI legislation
The clash between Anthropic and OpenAI highlights divisions in the AI industry over how to balance innovation with safety
The outcome could set precedent for whether states or the federal government lead on AI rules

Background

Anthropic was founded as a safety-conscious AI company and has positioned itself as more open to regulation than many rivals. It has been pitching an AI safety framework to policymakers for months. The leaked draft is a legislative version of that framework. OpenAI has pursued a different approach, supporting compromise bills written by lawmakers and seeking consistent rules across US states.

Key Perspectives

Anthropic: The company sees its draft as a necessary step to manage AI risks, with planned tweaks to account for rapid progress. It has long advocated for some form of federal oversight, including model-blocking powers. Safety advocates: They argue the leaked bill is too weak, pre-empts stronger state protections, and accepts too much risk. OpenAI: The company implicitly criticizes Anthropic for trying to write rules alone, asserting that democratic processes with elected officials should lead.

What to Watch

Whether lawmakers take up the Anthropic draft or other proposals
Reactions from other AI companies and state-level regulators
The eventual shape of any federal AI safety bill and its pre-emption of state laws

Sources

Anthropic’s leaked AI bill sparks backlash · Semafor

Checked

12 claims in this story checked against 1 source: 8 supported.

Show each claim

SupportedAnthropic’s leaked draft AI bill sparked backlash.Semafor: “Anthropic’s leaked AI bill sparks backlash.”
SupportedOpenAI warned that no single company should write the rules for the technology alone.Semafor: “No one company (including OpenAI) should write the rules alone for a technology with such profound implications for our economy, society and national security.”
SupportedAnthropic pitched a draft proposal to lawmakers on Capitol Hill, and it leaked late last week.Semafor: “The debate around the contours of an acceptable federal AI safety bill shifted again after a draft of a proposal Anthropic has been pitching to lawmakers on Capitol Hill leaked late last week.”
SupportedAI safety advocates argued that the draft did not go far enough, unfairly pre-empted state laws, and accepted too much possible risk.Semafor: “However, AI safety advocates are arguing it doesn’t go nearly far enough, unfairly pre-empts state laws, and accepts too much possible risk.”
SupportedA company spokesperson said the draft bill was a legislative version of an AI safety framework Anthropic had been pitching for months.Semafor: “That draft bill is a legislative version of an AI safety framework the company has been pitching for months, a company spokesperson told Semafor.”
SupportedAn Anthropic lobbyist posted about planned tweaks to reflect the pace of AI progress and the company’s position on pre-empting state laws.Semafor: “Anthropic, long seen as the safety-promoting frontier lab, found itself on the defensive, with one of its top lobbyists posting about tweaks it plans to make to the draft that more accurately reflect the dizzying pace of AI progress and the company’s stance on pre-empting state laws.”
Partly supportedThe debate over federal AI safety regulation intensified after the draft leaked.Semafor: “The debate around the contours of an acceptable federal AI safety bill shifted again after a draft of a proposal Anthropic has been pitching to lawmakers on Capitol Hill leaked late last week.”
SupportedOpenAI Chief Global Affairs Officer Chris Lehane wrote the quoted statement in a post Monday, implicitly directed at Anthropic.Semafor: “OpenAI quickly saw an opportunity to distinguish itself from its rival, with Chief Global Affairs Officer Chris Lehane writing in a post Monday implicitly directed at Anthropic: “Policy decisions must ultimately be made through our democratic processes with our elected officials leading. … No one company (including OpenAI) should write the rules alone for a technology with such profound implications for our economy, society and national security.””
Partly supportedThe risks of Anthropic’s advanced AI had become apparent, while its tensions with the US government had grown.Semafor: “As the frightening possibilities of its advanced AI have emerged — and as its fight with the US government has brewed — the company has started to seek regulation through methods that have promoted heated debate within the federal government”
SupportedAnthropic pitched terms that would allow officials to block the release of certain models.Semafor: “like pitching the terms that would allow officials to block the release of certain models”
Not in the sourcesOpenAI has generally been a more cautious Washington player.Semafor: “OpenAI has generally been a savvier, subtler Washington player”
Partly supportedOpenAI has sought to endorse compromise bills written by lawmakers and encouraged similar state frameworks.Semafor: “OpenAI has generally been a savvier, subtler Washington player, wanting to be seen as ready to endorse compromise federal AI bills written by lawmakers, as it pursues a strategy across the states of encouraging similar frameworks it finds acceptable.”