OpenAI has invited the U.S. government to enact “fair use” legislation for AI training, urging the adoption of a copyright policy enabling AI models to be trained on copyrighted content. This is among the suggestions submitted by OpenAI to the “AI Action Plan” of the U.S. government, which is a policy initiative seeking to transform American AI policy. OpenAI contends that the fair use doctrine has been pivotal in ensuring AI development within the U.S., facilitating a plethora of startups and research developments.
The company emphasizes that limiting AI training to content in the public domain would make it impossible to build effective AI systems. OpenAI has been faced with legal allegations over its use of copyrighted works to train AI models, and some owners of content have brought lawsuits against the company for copyright infringement. Despite these challenges, OpenAI asserts that applying copyrighted works to train AI models is fair use according to settled precedents of the law.
OpenAI’s suggestion also makes much of laissez-faire policy regarding regulating AI. It argues that an excessive level of regulation has the potential to smother innovation and dislodge the U.S. from the leading position it currently occupies in AI. This follows recent action taken by President Trump, like canceling the nation’s first executive order addressing AI and stressing the value in ensuring that America leads the way in AI.
In addition to encouraging fair use, OpenAI’s proposal also includes recommendations for voluntary collaboration between the federal government and private sector AI companies. The company suggests that such collaboration could facilitate faster implementation of AI technologies in government activities, potentially allowing the government to benefit from new AI services sooner.
OpenAI has also expressed its concerns regarding the competitive landscape, particularly the Chinese AI breakthroughs. The company stated in a paper that Chinese AI startups such as DeepSeek are significant threats to American AI companies. OpenAI’s action to shape AI policy comes at a time when the company is confronting a complex legal and PR landscape, such as ongoing litigation and competitor AI startups.
The legal uncertainty of AI training on copyrighted material remains a prominent issue. Even though OpenAI and other developers of AI argue that such utilization constitutes fair use, there is no established case law on the matter. Uncertainty is very expensive for AI developers, given that it leaves them vulnerable to legal issues and conformity issues.
In recent court proceedings, courts have begun to determine the fair use doctrine for AI training cases. One U.S. federal court denied Ross Intelligence’s claim of fair use, ruling that the firm’s use of copyrighted content for AI training purposes violated copyright law. This ruling may have minimal effect on large AI companies like OpenAI since their models are different.
OpenAI’s suggestion to establish fair use for AI training is part of a broader initiative to ensure U.S. AI policy encourages innovation and competitiveness. The suggestions the company has made would balance the need for regulatory certainty with the requirement to maintain the U.S.’s position at the forefront of AI development.
