Artificial intelligence is front and center in a murder appeal in Clayton County, Georgia. Aside from the sanctions imposed on various levels, the case is notable for bad citations that became part of the court’s official ruling, and which originated with the prosecution’s filings.

The same case first appeared here in Issue 3 (April 21), which was prior to Tuesday’s weigh-in from the state’s high court. Today’s issue includes Supplemental No. 01: Payne v. State, a full timeline and source map of how fabricated law moved from a prosecutor’s filing into a court order.

Professionals are worried their writing will get falsely accused of being AI-written, so they’re deliberately making it messier and more “human.” The fear has potentially flipped the “Turing test” on its head, creating a situation where humans feel they have to prove they are human, rather than a machine having to prove it is human. On the opposite end of the spectrum, U.S. AI firm Anthropic, maker of the popular Claude chatbot, was ordered to pay $1.5 billion for downloading books from pirate libraries prior to training its chatbot (if you’re wondering, the training on such books survived review).

Police in Oklahoma City are in the early stages of adopting AI. (The Washington Post)

Oregon’s single, statewide appellate court is raising alarms about the amount of fabricated authority in filings. (Oregon Courts)

Federal law enforcement released numbers on a month of work on child exploitation cases in Operation Iron Pursuit. (Fox News; DOJ)

Two U.S. cybersecurity professionals were sentenced to four years each for using ransonware on Americans while working in the industry. (DOJ)

Famed evolutionary biologist Richard Dawkins spent three days talking to new friends “Claudia” and “Claudius.” (UnHerd)