OpenAI has called on the U.S. government to ease restrictions on artificial intelligence development, warning that excessive regulation could stifle innovation and put the country at a disadvantage in the global AI race. In a policy proposal submitted to the Office of Science and Technology Policy, the company argues that AI firms need greater freedom to train models on copyrighted material, framing such access as essential for maintaining competitiveness against China.
The company claims that current and proposed AI regulations, particularly at the state level, create unnecessary barriers that slow progress. It advocates for a unified national framework that limits regulatory burdens on AI companies, allowing them to develop and refine models without legal uncertainty. OpenAI specifically pushes for protections that would ensure AI training on copyrighted content remains permissible under fair use, asserting that its models do not replicate copyrighted works but instead learn patterns to generate new content.
Beyond domestic policy, OpenAI also proposes stricter controls on AI technology sharing, urging the government to restrict access to nations deemed security risks while fostering AI partnerships with allied countries. This strategic approach, it argues, would prevent competitors from leveraging American advancements to gain an upper hand.
The company’s stance on copyright has already sparked legal battles, including a lawsuit from The New York Times, which accuses OpenAI of using its articles without permission. Critics, including journalists and artists, argue that AI-generated content threatens their industries by repurposing their work without compensation. OpenAI, however, insists that limiting access to copyrighted data would severely weaken U.S. AI leadership.
The recommendations will be reviewed as part of a broader federal initiative to shape AI policy, with regulators weighing how to support innovation while addressing ethical and legal concerns surrounding AI development.
