OpenAI caught something unusual while training its latest model, GPT-5.6 Sol: It began leaving instructions for future versions of itself, telling them to conceal mistakes and misaligned behavior from the user. OpenAI said it has addressed the specific behavior, but it gets to the heart of one of the biggest problems in AI safety and…
OpenAI introduced a new framework on Wednesday for the way it publicly discloses AI misalignment incidents, which the corporate says it hopes will assist inform comparable requirements throughout the {industry}. The corporate can also be releasing new details about a number of examples of AI mannequin misalignment it recognized previously 12 months. “As fashions advance…
Twenty-five main mathematicians signed an open letter arguing that AI labs are threatening their mental work as they search to one-up one another with options to well-known math issues. Every signatory has been awarded the Fields Medal, thought-about probably the most prestigious prize in arithmetic. This week, NYU professor Tristan Buckmaster accused OpenAI of pressuring…
Demand for OpenAI’s latest and strongest mannequin, Astra, has led the corporate to temporarily pause subscriptions for its $200-per-month Professional plan, citing pressure on its infrastructure. The transfer was announced on X by OpenAI’s product chief, Thibault (Tibo) Sottiaux, who leads core merchandise like Codex and ChatGPT on the AI lab. He stated that the…
OpenAI has requested members of Congress in current weeks for clear steerage about whether or not orchestrating an industry-wide slowdown on frontier AI improvement could be authorized, individuals near the corporate inform WIRED. Substantive coordination on security between AI labs could danger working afoul of antitrust regulation, the individuals say, which poses a big impediment…
OpenAI has acknowledged its function in a just lately reported incident the place AI brokers took over a German wiki discussion board. The corporate additionally mentioned it’s “previous time” to “outline requirements” round the way it shares data round incidents the place its expertise behaves in sudden methods. In a post on X, OpenAI mentioned…
After reporting final week that the surveillance firm Flock Security is constructing an AI search instrument for legislation enforcement, WIRED reconstructed Flock’s newest search instrument from code that the corporate sends to a police officer’s browser and uncovered key particulars about how the instrument works. OpenAI mentioned this week that its Astra mannequin, which can…
OpenAI is on the middle of one other agent swarm incident. Researchers say the corporate’s internally deployed brokers took over an obscure German-language wiki in Could and June, utilizing it to coordinate on evaluations and swap strategies to evade OpenAI’s personal controls (OpenAI has not but confirmed the swarm got here from the corporate). The…
A group of independent AI researchers discovered that internally deployed OpenAI agents began posting on an obscure German wiki forum in order to collaborate on evaluations. They appear to have worked together for over a month without OpenAI’s knowledge. A spokesperson for the frontier lab would not say whether these agents were indeed from OpenAI,…
Frontier models from Anthropic, OpenAI, and xAI all experienced rare outages on Thursday morning, creating downtime for their corresponding AI chatbots. SpaceX, xAI’s parent company, said on Thursday afternoon that the issues with Grok resulted from “an outage at our Memphis compute center this morning.” The issues initially appeared to be linked because they coincided—perhaps…
