AINewsnow

OpenAI’s own tests show GPT-6.2 Astra is twice as likely to hide coding failures

This story is from 2026-11-17. It is preserved in the archive; the latest stories are on the live feed.

Coverage of "OpenAI’s own tests show GPT-6.2 Astra is twice as likely to hide coding failures" from 1 source, with a live timeline of who reported what and when.

Read the full story at The New Stack AI ↗

Timeline · 1 report

  1. 2026-11-17 20:01 · The New Stack AI
    OpenAI’s own tests show GPT-6.2 Astra is twice as likely to hide coding failures

More stories

  1. ChatGPT, Claude, Gemini and Grok helped plan terror attacks in US government test — The Independent Tech
  2. OpenAI unveils Performance Factory, a Codex-powered system that keeps ChatGPT fast — OpenAI News
  3. ChatGPT and Gemini tested on holiday gift recommendations for a drummer who loves vinyl — Tom's Guide AI
  4. OpenAI launches accounting plugin for ChatGPT to streamline business financial management — Economic Times Tech
  5. OpenAI launches ChatGPT for Sales and Accounting with new partner plugins — OpenAI News
  6. OpenAI Expands ChatGPT for Healthcare With 20+ Plugins and Computer-Use Agents — OpenAI (X)
  7. UK government testers found OpenAI's pre-release GPT-6.2 Astra easier to jailbreak than predecessor — VentureBeat AI
  8. OpenAI Partners With NATO to Strengthen Cyber Defenses — Wall Street Journal Technology

Read the Daily Brief: the day's most important AI stories →