{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreifqlsntlynlfn5igtwblgpikhhla53jhjguqzsdzy4guh6y7tbgjq",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mo7mibbdkko2"
  },
  "path": "/t/gpt-5-5-seems-to-be-degraded/1381700?page=2#post_40",
  "publishedAt": "2026-06-14T00:51:48.000Z",
  "site": "https://community.openai.com",
  "textContent": "I have detected similar behaviour in the last days of using it - 14 iterations to fix simple frontend bug that Codex created itself. It took it 3 days for deduplicating data that it produced which I solved under 3 minutes myself at the end.\nCurrently I am facing serious security/privacy breach caused by Codex tools and Open AI support is not taking any real actions to investigate it though I explicitly escalated it several times. It seems that there is no support or interest in security issues by Open AI.\nI am not sure that I would recommend Open AI products to any one, on the contrary Open AI is not providing timely support. It is not acceptable for security/privacy issues about sensitive information no action to be taken for 96 hours now.",
  "title": "GPT 5.5 seems to be degraded"
}