OpenAI Questions if AI Industry Can Legally Slow Down

OpenAI has spent recent weeks pressing lawmakers on Capitol Hill for a straight answer to an odd question: could the AI industry legally agree to slow itself down? People close to the company tell WIRED that OpenAI wants clarity on whether coordinated action among AI labs to pump the brakes on frontier development would run into antitrust trouble. It's a real concern. Any formal agreement between competitors to limit output, even for safety reasons, invites scrutiny under laws like the Sherman Antitrust Act.

The question isn't hypothetical inside the company. OpenAI's chief scientist Jakub Pachocki published a blog post last weekend arguing that "coordinating to slow down future development" may be necessary to keep self-improving AI systems safe, and predicted that voluntary slowdowns will become routine until labs agree on shared safety standards. That kind of coordination, though, is exactly what antitrust law tends to punish when it happens between rival companies in the same market.

OpenAI didn't respond to a request for comment. But the timing lines up with movement in Congress, where a bipartisan group of lawmakers introduced legislation this summer that suggests some in Washington are already thinking about how to handle this.