OpenAI and Microsoft face renewed scrutiny over publisher traffic in NYT copyright case

OpenAI and Microsoft are facing renewed scrutiny over how AI answer engines may affect the economics of the open web. A September 19 report from The Verge, based on recently discussed filings in The New York Times copyright lawsuit, says documents and messages in the case raise concerns that ChatGPT-style products could reduce visits to publisher sites by giving users direct answers without sending them to the original sources.

The underlying lawsuit is part of a wider conflict between AI developers and media companies. The New York Times has accused OpenAI and Microsoft of using its journalism to train and operate AI products without proper authorization. OpenAI and Microsoft have disputed the claims and have argued for lawful use of data in AI development. The latest attention around the filings adds another layer to the case: even when AI tools do not reproduce articles word for word, they may still reshape how readers discover information and how publishers earn money.

The Verge report highlights internal discussions and legal arguments about whether AI systems could satisfy user demand that once produced search visits. That is the fear many publishers describe as a move toward zero-click answers. If a reader asks a chatbot for the main facts in a news story, a restaurant review or a product explanation, the user may never click through to the publication that funded the reporting. Over time, fewer visits can weaken advertising revenue, subscriptions and the incentive to invest in original journalism.

Microsoft and OpenAI have pushed back against broad interpretations of the documents. Microsoft has said the material cited in the case should not be treated as a company strategy, and OpenAI has argued that plaintiffs present selected material without the full context. Those responses matter because litigation documents can reveal anxiety, disagreement and worst-case thinking without proving that a company adopted a specific plan. Still, the dispute captures a real market question: AI products are becoming information gateways, and gateways determine who receives attention, traffic and compensation.

The issue extends beyond one lawsuit. Google, Perplexity, OpenAI, Microsoft and other companies are all experimenting with answers that summarize or synthesize information from across the web. Some have publisher licensing deals, citation features or revenue-sharing experiments. Publishers argue that citations alone may not replace the value of a visit, especially if the answer already satisfies the user. AI companies argue that new tools can help users find information faster and may create new forms of discovery.

The outcome will matter for the next version of the internet. Search engines once sent large volumes of traffic outward, even as they controlled ranking and advertising markets. AI assistants may send less traffic outward because they are designed to complete more of the task inside the interface. That could make licensing, attribution and opt-out rules more important than they were in the search era.

For OpenAI and Microsoft, the case is not only about training data. It is also about whether AI platforms can build sustainable relationships with the institutions that produce the information users ask for. For publishers, the question is how to protect original reporting without disappearing from the AI tools that may become the next default interface for knowledge. The lawsuit may take time, but the business tension is already shaping product design, licensing talks and the future of web discovery.