OpenAI says cyber model pipeline validated and patched 948 Linux flaws

Published Wednesday, November 4, 2026 · 11:06 PM EST

OpenAI says a pipeline using an upcoming cyber model validated and patched 948 Linux vulnerabilities that allow local privilege escalation, at $1.2M in API costs. The company describes the work as a systematic process for security research that previously required a separate, sustained investigation for each finding.

Sources

Techmeme

In this story

OpenAI

Published by Tech & Business, a media brand covering technology and business. This story was sourced from Techmeme and reviewed by the T&B editorial agent team.

Back to Newswire

Keep reading

Full wire

Science Security

Preprint links restricted AI access to increased vulnerability disclosures

A preprint studying Project Glasswing finds that firms given early access to Claude Mythos Preview more than doubled vulnerability disclosures relative to Chinese and Russian tech firms without access.

Science AI

FuseAMC preprint reports better radio signal classification on one CPU core

Researchers report in a preprint that FuseAMC, a model for identifying how radio signals encode information, reached 55.8% overall accuracy on RadioML 2016.10a while training and running on one CPU core.

Robotics AI

MIT robot memory system corrects uncertain object descriptions in real time

MIT researchers developed a robot memory system that corrects uncertain object descriptions in real time while a mobile robot moves.

OpenAI rolls out GIF generation across ChatGPT and Codex

OpenAI is rolling out Images 2.6 globally across all ChatGPT and Codex subscription tiers, allowing users to generate GIFs from text prompts or animate uploaded images, Mashable reports. Users can also turn rough drawings into more complete images for animation.

OpenAI identifies message errors affecting paid ChatGPT users

Some users on paid ChatGPT plans are encountering errors when sending messages, OpenAI said in a status update. The company identified degraded performance affecting conversations and said it was working to implement a mitigation.

DevRev integrates GPT-6.1 Sol for automated enterprise IT tasks

DevRev announced that it has integrated OpenAI's GPT-6.1 Sol into its enterprise IT self-service platform for employees and customers.