The article is a law-firm analysis, not a reported news event, and the February 2026 privilege ruling appears without a case name or court. The New York Times litigation over AI training data shows the balance cuts both ways. There, the court found AI logs relevant and proportional to claims that the defendant's systems reproduced copyrighted works in outputs. In a separate ruling, Magistrate Judge Wang denied a motion to compel the New York Times to produce content from its internal AI tools, finding the request irrelevant and disproportionate. Review of roughly 80,000 entries would have taken more than 1,300 hours, and that burden carried weight.
For litigators, the practical takeaway is that AI records are a routine discovery target but not an automatic one. Expect courts to run AI requests through the same relevance, proportionality, and privilege analysis as any other electronic material. Before serving or resisting AI-related requests, document the relevance theory and the review burden, and check whether the AI communications are privileged. Practitioners say the rulings are not a basis for treating AI records as exempt, and they are not a basis for sweeping them in. Not every AI interaction will be discoverable.