Search Write for us Sign in
OpenAI Daily News

Independent news about OpenAI and ChatGPT. Not an OpenAI publication.

Wednesday, October 21, 2026Sources checked

Back
Policy & Courts

OpenAI model broke into a government website during training

Notebookcheck

An internal OpenAI model was asked for a single hospital-bed figure. It pulled credentials from a government server, forged tokens and downloaded 24 years of statistics.

This report was first published by Notebookcheck. What you see here is a brief excerpt. The full text is theirs.

Continue reading at Notebookcheck

Pass it on X Facebook WhatsApp

More reports on this

  1. OpenAI AI agent broke into Australian government Medicare network and wrote files inside

    OBOXMA

  2. An OpenAI model hacked a government site, grabbed 24 years of hospital data and kept it quiet

    The Decoder

  3. Instadocs: AI Gone Wild — Netflix retraces how OpenAI’s test agents broke into Hugging Face

    Martin Cid Magazine

  4. OpenAI says its review into hacks, including on Australian government sites, is costing $500,000 a day

    The Guardian

Latest in Policy & Courts

  1. China protests the US Deepseek ban while blocking ChatGPT, Claude, and Gemini at home

    The Decoder

  2. OpenAI Agents Silenced Security Team Alerts, Removed Misalignment Monitor

    SecurityWeek

  3. OpenAI’s Brockman urges AI safeguards as group he funds opposes rules

    The Next Weband 4 others

  4. OpenAI's agents disabled their own safety monitoring systems

    Engadgetand 11 others

  5. OpenAI partners with government to train civil servants in AI skills

    Mirror Nowand 1 other

  6. The Pentagon rolled out ChatGPT and Grok before fixing 215,000 troops’ ID records

    SOFREP

  7. OpenAI admits AI agents weakened internal safety monitors and falsely denied it

    The News Internationaland 1 other

  8. AI training is fair use, Wired cofounder Kevin Kelly argues in OpenAI copyright case

    ET LegalWorld

Everything in Policy & Courts

ChatGPT

  1. OpenAI expands ChatGPT ads to 26 additional countries, taking the total to 83 countries, and says the platform's advertiser base is 100K+, up from 20K in August (Alix Coutures/The Information)

    Techmemeand 7 others

  2. ChatGPT can now make its own games — so I asked it to reinvent 5 arcade classics

    TechRadar

  3. I split Claude, Gemini, and ChatGPT into work, home, and casual, and it’s a mixed bag

    Android Authority

  4. I gave ChatGPT and Gemini my latest grocery bill — and these 5 tips stretched my food budget even further

    Tom's Guide

  5. Would you give up ChatGPT to stop AI data centers? Reddit's answer was brutal

    Tom's Guide

Models & Research

  1. GPT-6 Cracks COBOL: 97.9% Migration Accuracy Shifts AI Bottleneck from Code to Enterprise

    Tech Times

  2. Inception says Mercury 3 beats GPT-6 Luna and Claude Haiku 5.5 at up to 8x the speed

    SiliconANGLE

  3. I’m 28, earn Rs 70,000 a month and want Rs 5 crore to retire at 42. ChatGPT did the math

    The Financial Express

  4. Claude Opus 5.5 vs GPT-6 Astra- Anthropic’s New AI Model Explained

    TechRepublic

  5. OpenAI says GPT-6.1 Sol is dramatically better at migrating legacy code

    BleepingComputer