OpenAI submits proposals for the US AI Action Plan

OpenAI submitted a response to the White House Office of Science and Technology Policy’s request for information ahead of a national AI Action Plan, framing its proposals around competition with China. The document argued that restricting US AI developers’ ability to train on copyrighted material under fair use — while Chinese developers, OpenAI said, faced no equivalent restriction — would hand Beijing “an irreparable advantage” in frontier model development, and asked the federal government to secure a “freedom to learn” for American AI models alongside protections for rights-holders.

The submission also called for export policy that restricted advanced AI technology from reaching China while actively promoting its export to allied democratic countries, and pushed for federal preemption of the growing patchwork of state-level AI legislation. The framing was notably more combative than that of Anthropic’s submission to the same request a week earlier, which centred on national-security testing standards and power infrastructure rather than copyright or a race narrative against Beijing — a contrast commentators including Zvi Mowshowitz noted at the time.

The submission preceded the government’s own AI Action Plan, published that July, by more than four months; the eventual plan’s deregulatory tilt and emphasis on out-competing China echoed language close to what OpenAI had proposed, though the plan represented many inputs and cannot be attributed to any single submission.