Ameriška vlada podpira OpenAI v sporu o avtorskih pravicah z New York Timesom

The U.S. government has taken a firm stance in favor of OpenAI in its ongoing copyright dispute with the New York Times. In a “Statement of Interest” filed in federal court, the administration, under former President Donald Trump, argued that training artificial intelligence systems using protected content does not constitute a copyright violation. The move was justified on grounds of national security interests, with the government asserting that AI plays a critical role in intelligence analysis, weapon systems, and battlefield targeting. A ruling in favor of the Times could potentially harm U.S. technological competitiveness, according to the statement. The legal battle began on December 27, 2023, when the New York Times accused OpenAI and Microsoft of unlawfully using millions of articles to train their GPT models, which power services like ChatGPT and Copilot. The newspaper claimed this practice threatened its subscription, licensing, and advertising revenues. OpenAI, in turn, alleged that the Times had fabricated evidence by constructing prompts to support its claims. As the case progressed, the Times and other plaintiff publishers sought sanctions against OpenAI, alleging that the company had withheld relevant evidence and deleted billions of ChatGPT usage logs. This case is considered one of the central precedents in determining whether AI training qualifies as fair use under copyright law. This marks the first time the U.S. government has intervened in a major AI-related copyright case. While the administration is not a formal party to the lawsuit, its involvement signals a growing interest in shaping the legal landscape around AI technology. The filing, submitted in the U.S. District Court for the Southern District of New York (Case No. 1:25-cv-03483), argues that imposing licensing fees on AI training would disproportionately burden established media organizations. Instead, the government suggests that AI fosters a level playing field by enabling independent creators to compete with large publishing houses. The government’s arguments draw parallels between human learning and machine training. It cites examples such as author Joan Didion, who reportedly transcribed works by Ernest Hemingway to improve her own writing skills. By analogy, the government contends that AI can similarly enhance human creativity through training. Furthermore, the statement notes that the New York Times itself uses AI tools in its editorial process, suggesting the publication benefits from the very technology it seeks to regulate. This subtle point underscores the complexity of the issue, where both sides claim to be victims of the same technological shift. The document also advocates for a clear distinction between the training phase and the deployment of AI-generated content. According to the government, the use of protected materials during training does not make them publicly accessible, contrary to the Times's claims. The filing criticizes a previous ruling against Meta, arguing that it set a problematic precedent for how AI training should be evaluated. The government’s position emphasizes that while the Times has legitimate concerns, its approach risks undermining the broader potential of AI innovation. As the case moves forward, the implications extend beyond the immediate dispute. The outcome will influence how AI developers navigate copyright laws and whether they must secure licenses for training data. With the U.S. government backing OpenAI, the legal path appears to favor continued access to vast datasets, even if it challenges traditional notions of intellectual property. The resolution of this case could shape the future of AI development, affecting everything from content creation to national security applications.