This in-depth analysis by Chad A. Rutkowski of Baker & Hostetler unpacks the February 2025 Thomson Reuters v. Ross Intelligence decision, where the Delaware court granted summary judgment for Thomson Reuters—holding that Ross’s use of Westlaw headnotes to train its AI tool did not qualify as fair use. The court made clear that the training was commercial, non-transformative, and created a market substitute, emphasizing that even “intermediate” copying without substantial transformation can fail fair use protections. This ruling matters deeply to legal professionals: it sets a critical precedent that AI developers must carefully secure proper licenses for copyrighted training data, with fair use defenses facing steep judicial scrutiny. With broader implications for generative AI platforms like OpenAI or Meta, this piece is essential reading for IP counsel advising on AI-data compliance and litigation strategy—making it a must-click for forward-thinking practitioners.