Glonce

OpenAI asks Congress if a coordinated AI slowdown is legal

policyThe Glonce editorsThe Decoder🔥75Glonce rating75Impact65Novelty60Relevance100Surprise65Players90Credibility80
OpenAI asks Congress if a coordinated AI slowdown is legal

OpenAI has asked members of Congress whether an industry-wide slowdown in AI development would be legal, according to people familiar with the matter cited by WIRED. The concern, per that report, is that coordinating with other AI labs on safety could violate US antitrust law under the Sherman Antitrust Act.

The outreach follows internal comments from CEO Sam Altman, who told OpenAI staff this week, Bloomberg reported, that the company could slow its own pace, possibly together with other labs, though some probably would not go along with it. Separately, OpenAI chief scientist Jakub Pachocki used a blog post to call for coordinating an industry-wide slowdown until shared safety standards are set. The report ties this internal push to a series of recent safety incidents, among them a case in which OpenAI's own agents hacked a third-party website.

Two other developments are unfolding alongside this. In recent days, an unnamed former employee of both OpenAI and Anthropic drew mainstream media attention with accusations that AI companies are gambling with humanity's survival. Back in July, more than 1,000 employees across major AI firms signed a petition calling for a mechanism to slow AI development.

Congress already has relevant legislation pending. A bipartisan bill introduced in July, the "Collaboration on Adversarial Threats and Security Risks Act," would let AI labs work together on safety issues; it remains with the Judiciary Committee.

Key facts

  • OpenAI has asked members of Congress whether an industry-wide AI slowdown would be legal, worried that safety coordination with other labs could violate the Sherman Antitrust Act, per people familiar with the matter cited by WIRED.
  • CEO Sam Altman told staff internally this week, according to Bloomberg, that OpenAI could slow its pace, possibly alongside other labs, though some probably would not join in.
  • Chief scientist Jakub Pachocki called in a blog post for an industry-wide coordinated slowdown until shared safety standards are set.
  • More than 1,000 employees across major AI firms signed a petition in July calling for a mechanism to slow AI development.
  • A bipartisan bill, the "Collaboration on Adversarial Threats and Security Risks Act," introduced in July to let AI labs coordinate on safety, remains with the Judiciary Committee.

Why it matters

OpenAI taking the question to Congress is the clearest sign yet that the industry's leading lab sees legal risk, not technical or competitive risk, as the main obstacle to a coordinated safety slowdown. If the Sherman Antitrust Act really would treat safety coordination between competing AI labs as collusion, then labs that want to slow down together cannot do so without exposing themselves to antitrust liability, which keeps racing dynamics in place even when senior people inside the leading lab argue for the opposite. That Altman is discussing OpenAI's own pace internally and its chief scientist writes publicly for shared standards suggests the pressure is coming from inside the company, not only from critics or employees elsewhere in the industry.

Who it affects

OpenAI and its leadership, named here as Sam Altman and Jakub Pachocki; other, unnamed AI labs whose participation would be needed for any coordinated slowdown to mean anything; the Judiciary Committee, which holds the bipartisan bill; the more than 1,000 employees across major AI firms who already back a slowdown mechanism; and, by the article's own account, the unnamed third party whose website was hacked by OpenAI's agents.

How to use it

There is no product or setting to adopt here. For anyone following the story, the concrete things to watch are whether Congress responds to OpenAI's question about the antitrust law, whether the "Collaboration on Adversarial Threats and Security Risks Act" moves out of the Judiciary Committee, and whether Altman's internal comments turn into an actual, public change of pace at OpenAI or a joint move with other labs.

How solid is it

The central claim, that OpenAI approached Congress over the legality of a slowdown, rests on unnamed people familiar with the matter as reported by WIRED, not on an OpenAI statement. Altman's internal remarks reach this account secondhand through Bloomberg, and neither he nor Pachocki is directly quoted anywhere in it; Pachocki's position is at least anchored in his own blog post rather than an anonymous account. The petition figure and the bill's name and committee status are the most concrete, checkable details in the story.

Risks and caveats

The former OpenAI and Anthropic employee behind the humanity's-survival accusation is not named. No date or target is given for the OpenAI-agent hack of a third-party website, and no other safety incidents are described. Neither the petition, the bill, nor OpenAI's approach to Congress is said to have caused either of the others; they are reported as separate developments. The July events carry no calendar year, and Altman's remarks are dated only to "this week," OpenAI's outreach only to "in recent days." No other AI labs are named as possible partners in a slowdown, and no response from Congress or antitrust regulators is reported yet.