Sam Altman, chief executive of OpenAI, has defended the company’s use of copyrighted news articles and opinion content to train AI systems such as ChatGPT, citing the U.S. doctrine of fair use.
Speaking publicly, Altman stated that OpenAI’s approach is rooted in existing legal frameworks while the company simultaneously explores commercial partnerships with publishers and creators. He emphasized that when AI systems reference or quote news material, OpenAI aims to do so responsibly and in ways that could enable new business models for content creators.
“Our stance is that models can learn like people can learn,” Altman noted, adding that AI systems should not exploit content in ways humans legally cannot. He framed the company’s training methodology as an effort to allow AI to learn from publicly available material while remaining within intellectual property boundaries.
The remarks come at a time when multiple publishers have initiated legal proceedings against AI firms, alleging unauthorized use of copyrighted material in model training. These lawsuits argue that large language models rely on protected content without proper licensing or compensation.
OpenAI, however, maintains that its practices align with fair use principles under U.S. law. At the same time, the company has increasingly pursued licensing agreements and strategic partnerships with media organizations, signaling a parallel effort to formalize relationships with news publishers as regulatory and legal scrutiny intensifies.
