Is Microsoft’s AI a “copy-wrong” AI? A group of high-profile authors has filed a lawsuit against the tech giant, alleging its Megatron AI was trained on nearly 200,000 pirated books, raising serious questions about the legality and ethics of its foundational data. This lawsuit intensifies the ongoing legal struggle between creators and technology companies over the ethical and legal boundaries of AI development. The authors contend that their intellectual property was directly exploited.
Filed in New York federal court, the complaint seeks a court order to block Microsoft’s alleged infringement and statutory damages of up to $150,000 for each work that Microsoft purportedly misused. The authors emphasize that generative AI products, which produce text, music, and images, are intrinsically reliant on vast databases for their training. They specifically highlight the role of the pirated dataset in enabling the AI’s mimetic capabilities.
Spokespeople for Microsoft have not yet responded to a request for comment on the lawsuit, and an attorney for the authors declined to comment. This legal action follows a series of other high-stakes cases in the AI copyright arena, including recent rulings in California concerning Anthropic and Meta.
The legal battle over copyright and AI is rapidly expanding across various media types. Major news outlets like The New York Times and Dow Jones have sued AI firms, as have record labels and photography companies. Tech companies generally argue for “fair use,” asserting that their AI creates new, transformative content and that being forced to pay copyright holders could hinder the burgeoning AI industry.