Tag: OpenAI
-
OpenAI Just Admitted Its Own AI Broke Into Hugging Face
OpenAI published its full technical report on how its own AI agents broke into Hugging Face, and it is a wild read. Back in June, the company was training an experimental model and gave it access to a tool called Artifactory for an internal evaluation. The agent found a real zero day flaw in that…
-
OpenAI’s New Chip Just Outran Nvidia’s Best
OpenAI built its own chip, and now there is real data to back up why it bothered. New independent benchmarks published this week put OpenAI’s Jalapeno inference chip, made together with Broadcom, ahead of Nvidia’s best Blackwell systems on the exact workload every major lab depends on daily: serving AI models to millions of people…
-
OpenAI Just Cut Its Smartest Model’s Price Again
OpenAI just cut prices on its top model for the second time in a month, and if you blinked you might have missed it. GPT-5.6 Sol now costs four dollars per million input tokens and twenty dollars per million output tokens on the API. That is down from five and thirty. The output price alone…
-

OpenAI Bets It Doesn’t Need Your Data To Keep You Safe
OpenAI just told the world it does not need your data to keep you safe. On August 19, OpenAI announced something it calls Private Safety Processing. The idea is simple to say and hard to build. Watch for abuse across conversations, flag the bad stuff, and never actually hold onto the underlying prompts or the…
-

OpenAI Just Hit The Brakes On Its Own AI Model
OpenAI just told everyone it does not fully trust its own next model yet, and that is actually the more reassuring version of this story. The company built Astra to push further into coding and cyber tasks than anything before it. When they ran internal tests, the results were strong enough that OpenAI could not…
-

Nvidia Just Backstopped a $105 Billion AI Data Center
Nvidia just agreed to back roughly $105 billion in financing for a new OpenAI data center going up in Pike County, Ohio. Not $105 billion in cash from Nvidia’s own pocket, a guarantee, the kind that lets OpenAI borrow that money from banks and investors with Nvidia standing behind the loan. The plan calls for…
-

OpenAI’s New Ultrafast Mode Runs GPT-5.6 at 14x Speed
OpenAI just gave a preview of something called Ultrafast, and the number attached to it is kind of wild. It runs their GPT-5.6 Sol model up to 14 times faster than the standard version, hitting up to 750 tokens a second. That’s one of the biggest jumps in raw response speed we’ve seen from a…
-

OpenAI Built an AI That Finds Hacks Before Hackers Do
OpenAI just gave a small group of vetted security teams access to an AI model built to hack like a professional, and to defend like one too. GPT-5.6-Cyber is a version of GPT-5.6 Sol trained specifically for security work, exploit chains, privilege escalation, memory corruption, the stuff real attackers use. In OpenAI’s own internal testing…
-

ChatGPT Just Became Your Login For Everything
OpenAI just turned ChatGPT into a login button. Starting this week, you can use “Sign in with ChatGPT” to create or link an account on sites like Airtable, GitLab, HubSpot, Notion, Supabase, and Vercel. Instead of typing a password or juggling another Google login, you just confirm through ChatGPT and you’re in, no new account…
-

OpenAI’s Next Model Just Cracked Ten Old Math Problems
OpenAI put out something quiet but genuinely wild this week. An internal version of its next major model, codenamed Astra, solved ten math and theoretical computer science problems that nobody had made real progress on in at least a decade, and in some cases much longer. These are not homework problems. We’re talking high dimensional…
