Navigating the hazy intersection of generative artificial intelligence and intellectual property
The piece focuses on the “missing case law” problem: creators and researchers want predictable rules, but the precedent is still forming. It points to the U.S. Copyright Office’s partial revocation of “Zarya of the Dawn” (created entirely with generative AI) as a recent example of how strongly the system still ties copyright to human creative contribution. It also highlights patent policy moving in the same direction (e.g., courts treating “inventors” as natural persons), while still acknowledging that AI can meaningfully inform research without being the inventor.
On the infringement side, it calls out a set of high-profile lawsuits aimed at PerplexityAI — including a December 2025 complaint from The New York Times — alleging unauthorized distribution and training use of copyrighted work, plus downstream issues like hallucinated text attributed to publishers. The author’s practical takeaway is caution without paralysis: if you’re using LLMs or image models in creative or commercial work, assume the rules are still shifting, be transparent about your process, and keep humans firmly in the loop. Used thoughtfully, the article argues, AI can still be a tool for tedious tasks that frees up time for higher-value creative work.