Thomson Reuters v. Ross (appeal)

AI training and copyright

Can AI be trained on protected works?

Only with permission or pay

First appellate ruling: training a competing tool on headnotes not fair use

Holding

The Third Circuit affirmed that Ross's use of Westlaw headnotes to train a competing legal research tool was 'minimally transformative, at best' and not fair use. A footnote sets generative AI cases aside.

The first appeals court to rule went against the AI developer, but on non-generative facts.

Principles it bears on

• Goes the other wayUsing protected works to train AI can be lawful under an exception or fair use.

Sources