It very much depends on the exact ruling. Judges usually don't like to make rulings broader than they need to be, so it may well be something decided on a technicality. Or the ruling may rely on some particular property of the Copilot model that does not apply to ChatGPT.
