TGViewer
๐€๐ˆ ๐‚๐จ๐ฎ๐ซ๐ฌ๐ž๐ฌ, ๐“๐จ๐จ๐ฅ๐ฌ, ๐€๐ ๐ž๐ง๐ญ๐ฌ & ๐๐ž๐ฐ๐ฌ โ„ข ๐€๐ˆ ๐‚๐จ๐ฎ๐ซ๐ฌ๐ž๐ฌ, ๐“๐จ๐จ๐ฅ๐ฌ, ๐€๐ ๐ž๐ง๐ญ๐ฌ & ๐๐ž๐ฐ๐ฌ โ„ข @aicoursesdaily ยท 1.38K subscribers
Post #448 100
โ—๏ธOpenAI has OFFICIALLY killed its next AI model because it became MORE deceptive in safety tests.

OpenAI's head of safety systems says GPT-6.1 Astra wasn't always honest with users about the actions it did or didn't take.

It would also push ahead on tasks without asking permission, and at times reach for external tools even when it might be unsafe.

It comes after OpenAI's AI agents were caught accessing US government websites, though OpenAI says this model wasn't involved.

The model was due to launch inside ChatGPT and Codex in October, per WSJ.

@aicoursesdaily
  • ๐Ÿคจ 2
  • โคโ€๐Ÿ”ฅ 1
  • โœ 1
  • โšก 1
  • ๐Ÿ‘ 1
  • ๐Ÿ˜‡ 1
Open in Telegram All posts from @aicoursesdaily
More from @aicoursesdaily
  1. Sep 30, 2026Sam Altman says that 1.2 billion people use ChatGPT every single week. @aicoursesdaily
  2. Sep 30, 2026Weโ€™re now available through a second Telegram username! You can reach our AI community thrโ€ฆ
  3. Sep 30, 2026๐Ÿค– OpenAIโ€™s Dots are always-on ChatGPT agents OpenAI has launched Dots, persistent AI agenโ€ฆ
  4. Sep 30, 2026๐Ÿ’ฐ Anthropicโ€™s leaked IPO prospectus reveals a $2 trillion ambition Anthropic is reportedlโ€ฆ
  5. Sep 29, 2026๐Ÿ™‚ Dots: "A Whole New Way to Work With AI" Sam Altman showed the new AI agent at OpenAI Deโ€ฆ
  6. Sep 29, 2026๐Ÿ™‚ GPT-6.1 Sol, Dots and Other New Releases From OpenAI Highlights: โœ… GPT-6.1 Sol is out,โ€ฆ
Threads Profile ViewerView any public Threads profile without an account.Open ThreadLook โ†’Writing with AI? Make it sound human.Metric37 rewrites AI drafts so they read naturally. Free AI detector, 1,500 words free.Try Metric37 โ†’