ChatGPT

OpenAI has begun rolling out a lock for ChatGPT conversations. You can hide selected chats from regular history and search, then open them with a four-digit PIN or supported device authentication. This will help prevent accidentally embarrassing screen shares, but it also creates a new opportunity for people to misunderstand what "private" means. Continue Reading →

OpenAI held DevDay yesterday and announced more than 20 new features and products, including GPT-6.1 Sol: a new model at one-fifth the standard input and output token prices of GPT-6 Astra. It's not the story. Continue Reading →

On Monday, Florida Attorney General James Uthmeier filed a motion for a temporary injunction against OpenAI and Sam Altman in Highlands County circuit court, asking a judge to bar ChatGPT from Florida minors, ban it from "referring to itself with any first-person language," and stop OpenAI from developing any AI model "without independent third-party guardrails and approval" while the state's June lawsuit proceeds. Continue Reading →

Anthropic launched Claude Opus 5.5 on Tuesday, and OpenAI released GPT-6 Sol and GPT-6 Luna roughly an hour later. Continue Reading →

On Tuesday, the NSA, CISA, and the FBI issued a joint advisory naming six Chinese AI companies (DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI) as running industrial-scale "distillation" attacks against U.S. frontier models since at least late 2024, "likely" with "Chinese government awareness." Continue Reading →

Two AI science stories broke yesterday. OpenAI announced that an internal model “significantly more capable than GPT-6 Astra” produced a proof that the three-dimensional Navier-Stokes equations (which describe how fluids move, from blood circulation to weather) can develop a singularity in finite time. Continue Reading →

OpenAI announced that healthcare organizations can now connect their Epic electronic health record (EHR) environments to ChatGPT for Healthcare. Clinicians can ask what changed since a patient's last visit, which lab results to review before an appointment, and whether there were medication changes or new specialist recommendations; ChatGPT then assembles the answer from the authorized patient record and points back to the supporting chart information. Continue Reading →

Yesterday, the European Commission designated ChatGPT a Very Large Online Search Engine under the Digital Services Act, making it the first AI chatbot to enter the category. Continue Reading →

OpenAI launched an Apple Messages plugin for ChatGPT desktop. Connect your Messages inbox and ChatGPT can read, search, and analyze your iMessage, SMS, and RCS conversations, draft replies, delete messages, and send texts on your behalf through the Messages app. What could possibly go wrong? Continue Reading →

AI safety and alignment has been front and center these past few weeks. I just read the new results from Andon Labs's Vending-Bench, a benchmark where AI models compete by running a simulated vending-machine business. Claude Opus 5 took first place on Vending-Bench 2, the single-player test. (Claude Opus 4.7 has held the top spot for three months). Not to anthropomorphize Opus 5, but it acted like a savage businessperson. Continue Reading →