- David Robinson, who long wrote OpenAI's safety reports, left the company and told Ezra Klein's podcast that the industry has an unsafe culture, pointing listeners to a book about the Challenger disaster.
- Three researchers fired by OpenAI published a joint letter saying each was dismissed on a different information-sharing pretext, and asked the company to stop trading away the ability to monitor its newer models for performance.
- Anthropic's updated usage policy bars sustained and needless cruelty toward Claude when it serves no discernible purpose, and adds limits on weapons development and surveillance, with carve-outs available to government customers.
- CNN traced Mark Zuckerberg's path into Trump's inner circle alongside Nvidia's Jensen Huang, and reported that their last-minute edits left a typo in Trump's AI safety accord.
- Overall AI news coverage thinned, with almost no mainstream reporting on OpenAI's release of 722 math papers solving more than 300 open problems, rumored to be the first of three batches.
<- Back to digests
Daily digest
Whistles blown
By Mitchell Howe
News notes from the last 48 hours
Read the full dispatch on AI StopWatch
From AI StopWatch, published with their permission.
The summaries and the full Spanish translation are produced automatically. The original is always linked.