When the New York Times filed a copyright infringement lawsuit against OpenAI — the vendor of generative AI systems ChatGPT and DALL-E3 — it dragged complaints many authors, artists, photographers, and even programmers had out onto a large stage. The results could ultimately affect OpenAI's ability to operate and make money as well as the results its customers seek, including those in commercial real estate.

At the center of the conflict are control and compensation. Those whose work is creative in nature, and the companies that hire them and license their materials, don't have a say in how their material is used when companies scan vast amounts of text, imagery, or code to expose generative artificial intelligence systems for training purposes. The AI systems break the material down to individual parts and statistically track how one part — a word or image segment or bit of code — relates to what might follow. When someone provides a prompt, the system looks at the connections of content bits within the context and assembles the output.

The copyright issues on this are complex and many have never been litigated, making definitive legal interpretation unavailable. There is precedent that copying material into memory without permission at least is allowable. The AI software might be considered a derivative work, although that doesn't necessarily rule out the concept of fair use.