As Judge Chhabria indicated two weeks ago at a hearing in Kadrey v. Meta, he dismissed all of Kadrey’s claims except the copyright claim alleging infringement based on the Meta’s training of its large language model LLaMA.
We asked ChatGPT to summarize the court’s order and it did a decent job. We made some minor edits to the summary below.
Judge ChHabria’s reasons to dismiss Kadrey’s claims v. Meta
| Kadrey’s Claim v. Meta | Court’s Reason for Dismissal | Dismissal Status |
| Count 1 LLaMA Language Models as Infringing Derivative Works (*motion didn’t apply to training copies) | Found nonsensical to consider LLaMA models as a recasting or adaptation of the plaintiffs’ books; not qualifying as derivative works. | Dismissed with leave to amend within 21 days |
| Count 2 LLaMA Outputs as Infringing Derivative Works and Vicarious Copyright Infringement | Lack of specific allegations regarding the content of any LLaMA output that could be considered as infringing derivative works. | Dismissed with leave to amend within 21 days |
| Count 3 Digital Millennium Copyright Act Section 1202(b) Claims | Absence of facts supporting allegations that LLaMA distributed the plaintiffs’ books without their Copyright Management Information (CMI). | Dismissed with leave to amend within 21 days |
| Count 4 Unfair Competition Law (UCL) | Preemption by the surviving direct copyright infringement claim and lack of adequate allegations of fraud or unfairness. | Dismissed with leave to amend within 21 days |
| Counts 5-6 Claims for Unjust Enrichment Negligence | Unjust enrichment claim preempted by the Copyright Act; negligence claim preempted by the Copyright Act or, alternatively, barred due to the economic loss doctrine in tort law. | Unjust enrichment dismissed with leave to amend within 21 days; negligence dismissed with prejudice |