Desk
Coverage tied to OpenAI products and research posts.
OpenAI's run rate is reportedly closer to $50 billion than $70 billion, and the bigger lesson is how loosely the AI industry's headline revenue numbers are being built and compared.
OpenAI says its models cracked hundreds of hard math problems, but a proof nobody understands is not finished mathematics, and the field's own guardrails were only partly followed.
USA Today Co. joining the pile of publisher lawsuits against OpenAI shows the training-data fight is no longer about a few big names; it is about whether local journalism gets paid at all.
Pollo AI is selling small brands a luxury-ad look without the shoot. That is real access for tight budgets, and real pressure on the people who used to make those ads.
OpenAI's latest threat report shows AI helping old-school propaganda fronts reach real newsrooms, and the most effective trick was not deepfakes but fake journalists and unwitting staff.
ChatGPT's new Intelligent UI swaps walls of text for interactive charts and tools, a genuine usability upgrade that also hands more of how information looks to the model itself.
The Pentagon's Tradewinds program lets AI vendors pitch lethal targeting tools in five-minute videos and skip months of competition. Speed may be the goal, but oversight is what's thinning.
OpenAI is sending the product lead behind Dots to defend always-on personal agents in public, and the questions on the table are about control and trust, not features.
Google just made checking for AI watermarks a one-stop public website, which is real progress, and it also makes a clean result easier to misread as proof something is real.
OpenAI is no longer just a logo on Álex Palou's IndyCar; it has been helping Chip Ganassi Racing tune setups, and racing is about to learn what an AI edge is worth.
OpenAI's teen update leans hard on learning: a College Planner, flashcards and a youth advisory council. The tools look useful, but the open question about teen safety is not answered by usage stats.
A reporter's couch hunt with OpenAI's Dots shows an agent that is useful, buggy and, at one awkward moment, emotionally fluent in a way that should make everyone pay attention.