OpenAI has asked the Trump administration to declare training AI on copyrighted material a “fair use,” arguing that unrestricted access to data is key to the U.S.’s global leadership in artificial intelligence.
Image source: OpenAI
As Ars Technica writes, OpenAI hopes that Trump’s AI action plan, which will be presented in July, will resolve copyright disputes in favor of tech companies. This, according to the company, will ensure the US leadership in the global AI race, allowing American developers to freely use data to train and educate models. Otherwise, OpenAI argues, China will be able to gain an advantage due to freer access to this same data.
US courts are currently weighing whether it is legal to use copyrighted material to train AI, as copyright holders argue that such technology could undermine their commercial value and hurt creativity. AI companies face dozens of lawsuits in which plaintiffs claim that models are reproducing their work without permission. OpenAI, for its part, argues that AI does not copy content, but learns from it, analyzing it and transforming it into new products.
The court had previously ruled in favor of copyright holders in the Thomson Reuters case, ruling that training AI on their materials was not fair use because the models’ output potentially replaced the original product. That set a precedent that could influence other cases, including the New York Times’ major lawsuit against OpenAI and Microsoft.
Recognizing this risk, the company hopes that the Trump administration will step in and change its approach to copyright regulation, ensuring that the AI industry can “learn freely.” OpenAI argues that banning copyrighted materials from being used to train AI would put American companies at a disadvantage compared to China. “If Chinese developers have unrestricted access to data, and American companies are left without fair use rights, the AI race is effectively over,” OpenAI said.
Sam Altman. Image source: OpenAI Forum
In addition, OpenAI opposes many legislative initiatives proposed by individual US states that would limit the development of AI, citing the experience of the European Union. For example, in 2025 alone, 832 laws regulating AI have already been proposed in the US.
OpenAI warns that such measures could weaken U.S. economic competitiveness and threaten national security by allowing Chinese companies to operate without similar restrictions. Potential threats include manipulation of AI models, cybercrime, and identity fraud.
To protect the American AI industry, OpenAI called on Trump to intervene and pass federal legislation that would replace fragmented state laws and establish partnerships between the government and tech companies. Under such collaborations, AI companies would share knowledge and data in exchange for legal protections and protection from lawsuits.
OpenAI also suggested that the US should be more involved in shaping international copyright policy to prevent other countries from imposing restrictions on US companies, arguing that China is rapidly increasing its technological capabilities and that the US needs to take urgent action to maintain its leadership in AI.