Flaw in OpenAI's Codex Could Have Allowed Malicious Code via Fake Bug Report

Researchers found that a single fake bug report could have let attackers inject harmful code into OpenAI's Codex, which was downloaded 96.8 million times in one month. OpenAI has reportedly addressed the issue.

Read the full story at DailyAI ↗

Timeline · 1 report

2026-11-12 09:15 · DailyAI
A Flaw in OpenAI’s Codex Put 97 Million Monthly Downloads at Risk

More stories

Developers converge on shared tooling for Claude Code, Cursor, and Codex workflows — ZDNET AI
OpenAI launches ModRetro Chromatic Codex edition for building handheld games — OpenAI (X)
OpenAI Codex and Claude Code $20 plans compared on usage limits — Tom's Guide AI
OpenAI Showcases Voice Steering, Cloud Handoff for Codex at AI Engineer Code Summit — Latent Space
Google expands Gemini 4 Carbon access to Pro plan users in AI Mode — Google DeepMind Blog
Holiday Extras uses OpenAI's Codex to boost app development and bookings — OpenAI News
Anthropic launches Claude Fable 5.2 with lower API pricing — Anthropic (X)
Developers build a wave of MCP servers to extend Claude, Cursor and Codex — DEV Community — AI

Get the daily brief of stories like this at 6:30 every morning →