As a high-profile copyright lawsuit between major news publishers and AI companies unfolds, Microsoft is working to shield its recently launched consumer-facing Copilot AI product—and the team behind it—from the legal spotlight. In new court filings, the company argues that its updated consumer Copilot, a product of its Redmond-based AI division under Mustafa Suleyman, should be excluded from the lawsuit’s discovery process.
The lawsuit, originally filed in December 2023 in federal court in New York, accuses Microsoft and OpenAI of training their AI models on millions of copyrighted news articles without permission. The plaintiffs argue that by using their journalism to generate AI responses, the tech companies are undercutting the market for original reporting and diverting readers away from publishers’ websites.
The case is being led by a coalition of prominent news organizations, including the New York Times, but it doesn’t stand alone. Joining the legal effort are Daily News LP, the parent company of the New York Daily News, and The Center for Investigative Reporting, an independent nonprofit known for producing in-depth accountability journalism. Together, the group contends that unauthorized use of their work undermines both the economics and ethics of journalism in the digital era.
In the latest development, Microsoft is opposing demands by the plaintiffs to include documents and data related to its “new” Copilot AI, which was launched in October 2024 as a consumer product distinct from Microsoft 365 Copilot, the business version of the AI assistant. Microsoft says the product is not part of the original complaint, emphasizing that it was developed months later, on a different platform, with a newly hired team, and based on a separate architecture.
In legal terms, Microsoft’s argument centers on the scope of discovery, the process by which both sides in a lawsuit must share relevant evidence. The company contends that including the consumer Copilot would force them to identify and produce records from “hundreds of potential new custodians,” referring to the engineers and staff who built the product. Many of those developers, the filing notes, came to Microsoft from Inflection AI, the startup co-founded by Suleyman before he joined Microsoft in March 2024.
However, attorneys for the plaintiffs say these distinctions are largely superficial. They argue that the new Copilot is still powered by GPT-4o, an OpenAI model that’s already a focal point in the lawsuit. To them, the technology may have a new interface or name, but the core functionality and alleged harms remain the same: using AI trained on copyrighted material to generate news-like content, which could ultimately pull readers—and advertising dollars—away from the original sources.
The plaintiffs maintain that denying access to documents related to the consumer Copilot limits their ability to gather evidence and challenge Microsoft’s anticipated fair use defense—a central point in the legal battle.
So far, the court has not issued a final ruling on whether the AI models in question violate copyright law, and discovery disputes like this one are common in cases involving emerging technologies. The outcome, however, could shape future standards for how AI companies use public or commercial data to train their systems, and what legal responsibilities they carry when their tools generate content derived from that data.
Microsoft’s consumer Copilot product has quickly become a high-visibility feature across its platforms, offering users personalized news summaries, natural voice responses, and a browsing companion that can summarize or explain content in real time. Though Suleyman is not named directly in court filings, the legal language makes it clear that his team’s work is central to the product at stake.
In the meantime, both sides continue to spar over what data must be handed over, who should be deposed, and how the line should be drawn between legacy models and their evolving counterparts. For Microsoft, the stakes are as much about legal precedent as they are about protecting its most consumer-facing AI innovations—including those developed just up the road in Redmond.
