OpenAI seeks US guidance on legality of coordinated AI slowdown
WIRED reports that OpenAI has asked lawmakers whether frontier AI companies could coordinate a voluntary slowdown without breaching antitrust law.
OpenAI has approached members of the US Congress for guidance on whether AI companies could legally coordinate a slowdown in frontier AI development, according to WIRED, as concerns grow that antitrust rules could obstruct safety cooperation.
OpenAI chief scientist Jakub Pachocki has argued that voluntary slowdowns may be needed until shared safety standards are established. The company has not said that it has agreed to or implemented such a slowdown, and the specific guidance it sought remains unclear.
Legal scholars say a coordinated pause could be viewed as an agreement to restrict output under US antitrust law, depending on its structure and conduct. Nicholas Felstead, an assistant director at the Australian Competition and Consumer Commission, has previously warned that legal uncertainty could deter safety collaboration even where cooperation might ultimately withstand scrutiny.
Congress has taken an early step towards addressing the issue. A bipartisan, bicameral bill introduced in July would allow AI laboratories to coordinate on security and safety work without risking violations of antitrust statutes. The House version has been referred to the Judiciary Committee but has not advanced further.
The proposed cooperation also faces commercial and strategic barriers. AI companies are competing for a nascent frontier-model market, some executives view staying ahead of China as a national-security priority, and developers disagree over how safe AI should be built. Those differences may complicate any industry-wide agreement, alongside the legal questions.
