Andy Stone told me that any kids uncomfortable with chatbot outputs can report the chatbot the same way they would a predator. But Meta child safety whistleblower Arturo Bejar says there are easy ways Meta won't implement to actually monitor how many kids receive creepy outputs.

I also went through legal docs and old tweets, so you don't have to, which may be helpful to anyone trying to read through the lines of Elon Musk and Sam Altman's bubbling AI feud.

Seems like a good time to re-up this story where I talked to Cory Doctorow and one of Meta's only early Facebook rivals about Meta's (failing) plan to use AI friends to keep users locked into its apps: arstechnica.com/tech-policy/...

Wherein I explain that if YouTube's AI gets a user's age wrong, it will be their burden to prove how old they are, with very few privacy protections for the data collected -- and that is creepy and inhibiting and risky arstechnica.com/tech-policy/...

I had planned to do a live tariff talk tomorrow with CTA, but CTA's expert, for unforeseeable reasons, had to pull out, so now I'm afraid the world must wait until whenever chip tariffs drop (maybe next week) for me to do another deep dive. Very sorry for the short notice!

Join me Thursday to talk chip tariffs:

Dozens of xAI employees expressed concerns—and many objected—when asked to record videos of their facial expressions, according to a new report.

Moraine Hills State Park lived up to the hype and gifted me and my dog our first Illinois beaver sighting for her 8th birthday.

Some chipmakers are already feeling the pain of Trump's trade war, despite a 90-day pause on reciprocal tariffs and a tenuous exception for semiconductors and other electronics.

The recent ruling affirmed that AI art can’t be copyrighted—and machines can’t be authors (for now, at least).

The law could lead to some huge penalties for any Big Tech companies deemed non-compliant.