Judge rules witness names should stay in court records

The court in its judgment stated the sections of the redaction statute requiring removal of victim and witness names, addresses and contact information from court pleadings to violate the First and Fourteenth Amendments of the United States Constitution and the Open Courts requirement of the Missouri Constitution.

Pleasantly, I needed to rewrite this column because right before Christmas, a Cole County judge declared unconstitutional the statute requiring redaction of witness names in court records. The lawsuit – Gross v. Parson, 24AC-CC04658 – had been pending for about six months, and oral argument had been held in early December.
The court in its judgment stated the sections of the redaction statute requiring removal of victim and witness names, addresses and contact information from court pleadings to “violate the First and Fourteenth Amendments of the United States Constitution and the Open Courts requirement of Article I, § 14 of the Missouri Constitution.”
Quite a gift for those of us covering the news!
Normally, such decisions are appealed. That’s likely to be the case here too. But it is also possible the Missouri Attorney General’s office decides to not appeal this outcome. The office took an intriguing position in its argument regarding the redaction statute that it “should be interpreted as mandating redaction only when victim and witness names are required by [law] to be kept confidential.” Such a position essentially concedes that the redaction statute, as written, would offend the Missouri Constitution.
The state likely has 30 days to appeal this decision, so mark your calendars for the end of January.

****

I was in Willow Springs in October for the Ozark Press Association convention when I asked the attendees if they had ever attempted this experiment: Querying an AI product to write a news story in the style of their newspaper. Interestingly, someone raised a hand – and recounted how the AI product had reproduced a copy of one of the newspaper’s own articles.
This is the gist behind several of a trio of lawsuits brought by the Center for Investigative Reporting, the New York Times and the Daily News and other media companies against various AI companies in federal court in New York.
AI products are based upon large language models (LLMs). LLMs work by predicting language responses, and those predictions are based on the potentially billions of language examples used to “train” the model.
These lawsuits allege that companies fed LLMs vast quantities of their content without permission, and then often spit out identical copies or nearly identical copies when queried by a user. This is a phenomenon called “memorization.” Copyright violations are the chief cause of action in these lawsuits, but the media enterprises are also suing under common law for misappropriation of their content.
All of these cases are currently in early-stage litigation. The defendants, which include companies like Microsoft and OpenAI, have sought to have various portions of the plaintiffs’ cases dismissed. But it seems likely at least some claims will survive and that the cases will move forward.
Interestingly, last year OpenAI struck a deal with the Associated Press to license use of its news articles going back to 1985. It has similarly signed deals with The Atlantic, Vox Media, and News Corp. publications such as The Wall Street Journal.
These lawsuits are not “class actions” – these papers do not represent any of your interests. But they will set important precedent. You might take a moment and ask the AI product of your choice to write a news story about the town your paper covers. It would be interesting to see how many Missouri Press Association members find that their content has been “memorized.” Let me know what you find.