It's my dear hope that this post will be one of the few that appeals to muggles. Please, if you see one, thrust your phone at them proudly and demand they read it. Watch their terrified eyes scan over every word. Do not allow them to flee. Repeat. Repeat. Repeat!
Judge Bibas gave us this gift of a general interest (comparatively) post with his decision yesterday in Thomson Reuters Enterprise Centre GmbH v. ROSS Intelligence Inc., C.A. No. 20-613-SB (D. Del. Sept. 25, 2023) (Mem. Op.).
The case deals with the exceptionally buzzy issue of AI data scrubbing. Thompson Reuters runs the hugely popular Westlaw legal research platform. As part of that service they provide they provide vaguely useful "headnotes" describing the holdings of the cases.
(Eds. Note - Westlaw, if you are reading this, we can be bought. Every 10% you discount our service will result in an improvement of the adjective above. We can go all the way from "vaguely", to "quite", to "masterfully." It's about time this blog started paying the bills).
Ross is attempting to start some sort of AI-driven competitor where you just type in a question and receive a plain language legal answer. No Booleans, no problem.
To accomplish this, they hired a third party to turn many thousands of Westlaw headnotes into a series of legal memos. Some of these memos amounted to just copying the headnotes. The interesting dispute (actually the opinion has quite a lot of interesting bits about copyright, if that's your thing) was whether feeding these copied memos into an AI's insatiable pie-hole was fair use.
To resolve the issue, Judge Bibas analyzed whether the use was transformative, finding that it depended on exactly how the information was used:
So whether the intermediate copying caselaw tells us that Ross’s use was transformative depends on the precise nature of Ross’s actions. It was transformative intermediate copying if Ross’s AI only studied the language patterns in the headnotes to learn how to produce judicial opinion quotes. But if Thomson Reuters is right that Ross used the untransformed text of headnotes to get its AI to replicate and reproduce the creative drafting done by Westlaw’s attorney-editors, then Ross’s comparisons to cases like Sega and Sony are not apt. Again, this is a material question of fact that the jury needs to decide
Id. at 19-20.
As you'll guess from the above, the Court denied the parties cross-motions for summary judgment on the fair use issue. This'll be a fun one to watch if ti gets to trial -- we'll keep you in the loop.
If you enjoyed this post, consider subscribing to receive free e-mail updates about new posts.
