Judge dismisses bloated privacy class action against OpenAI and Microsoft
In Cousart v. OpenAI, Judge Vince Chhabria gave plaintiffs 21 days to refile a shorter complaint, calling the 204-page document nearly impossible to parse.
- Courts & copyright
- Minor
A federal judge in the Northern District of California dismissed a proposed class action accusing OpenAI and Microsoft of building ChatGPT and other generative systems on personal data scraped from the internet without consent. The case, Cousart v. OpenAI LP, had been filed the previous year and alleged violations of privacy, property and consumer-protection law across a complaint that ran to 204 pages.
US District Judge Vince Chhabria’s order was two pages long and procedural rather than substantive: he did not rule on whether the underlying claims had merit, only that the complaint itself was unworkable. He wrote that it was “not only excessive in length” but “contains swaths of unnecessary and distracting allegations,” making it “nearly impossible” to assess whether the plaintiffs’ legal theories were adequately pleaded. Plaintiffs were given 21 days to file a shorter, amended complaint.
The ruling was widely read as a rebuke of drafting strategy rather than a verdict on the AI-training data question at the heart of dozens of similar suits against major labs. It left the broader legal question — whether training a model on scraped personal data is itself an actionable harm — untouched, to be litigated again once, and if, a narrower complaint was filed.