When The New York Times filed a copyright infringement lawsuit against artificial intelligence startup OpenAI in December, it seemed like an inevitable face-off between old and new media. The Times accused OpenAI of scraping articles without permission to train its buzzy chatbot ChatGPT. In a strongly worded blog post on Monday, OpenAI claimed the Times‘ complaint does not tell the whole story and that any verbatim text reproduced by its AI systems are “rare bugs.”

At the heart of the conflict is how AI systems like ChatGPT work. The technology relies on ingesting massive volumes of text data — news articles, Wikipedia pages, forum discussions and more — to understand language and generate human-like responses. OpenAI maintains it used Times articles legally under fair use provisions to create general-purpose conversational AI.

But the Times highlighted several examples where ChatGPT spits back paragraphs lifted word-for-word from its articles, claiming this crosses the line into copyright violation. OpenAI countered that the newspaper may have intentionally tried to trigger such errors by manipulating prompts. It also accused the Times of “cherry-picked” instances out of potentially thousands of queries.

“The blog concedes that OpenAI used The Times’s work, along with the work of many others, to build ChatGPT,” the Times‘ lawyer Ian Crosby responded. “Defendants seek to free-ride on The Times’s massive investment in its journalism by using it to build substitutive products without permission or payment.”

The debate crystallizes larger concerns around using copyrighted digital content to power AI applications that could disrupt industries like journalism. Generative language models have already transformed creative sectors like art, music and literature — where does news fit in?

For the Times, the answer is clear: Compensate us for using our reporting or stop doing it. OpenAI claims its systems benefit public discourse by answering questions and summarizing current events. At a deeper level, the showdown raises philosophical questions about AI authorship and ownership.

As this battle between media and tech heavyweights picks up steam, one thing is certain: Precedents set by the case will reshape the ethics and business models around AI for years to come. Both sides have dug in their heels, believing the future is on the line.