{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreifsmawuzgha6xngv5bsfr3f5fe6dthi26fdzf57vt7ynb4re6ksne",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mebymfy2l3v2"
  },
  "path": "/t/feature-request-add-explicit-reference-declaration-for-external-sources/1373591#post_1",
  "publishedAt": "2026-02-07T16:02:16.000Z",
  "site": "https://community.openai.com",
  "textContent": "[Proposal] If AI is to become a real industry, trust must come before capability\n\nI am an individual user, but if I were a business decision maker considering ChatGPT for professional use, I would likely cancel the contract within the first month under the current design, possibly even request a refund.\n\nThe reason is simple:\n\nI cannot tell what the AI is actually referencing.\n\nToday, AI is already:\n\n- fast\n\n- capable\n\n- multifunctional\n\nHowever, in business contexts, the most important requirement is not intelligence, but trust.\n\nSpecifically, the ability to explain decisions and place responsibility.\n\nRegarding external file integrations such as Google Drive, there is a major gap:\n\neven when integrations exist, users cannot clearly see\n\n- which files were referenced\n\n- whether they were referenced at all\n\nThis makes review, auditing, and accountability extremely difficult.\n\nMy proposal is intentionally minimal.\n\nAt the very beginning of each response, simply display a short declaration such as:\n\n“Referenced files:\n\n- File A (last updated: YYYY-MM-DD)\n\n- File B (last updated: YYYY-MM-DD)”\n\nThis single line would immediately allow users to:\n\n- verify assumptions\n\n- detect reference errors\n\n- review outputs responsibly\n\n- treat the AI as a tool rather than a gamble\n\nSociety already accepts that AI can make mistakes.\n\nWhat matters is whether humans can detect and correct those mistakes.\n\nWithout visible references, users are forced to rely on trust alone.\n\nWith visible references, trust becomes structural.\n\nFocusing only on short-term feature expansion risks quietly losing users.\n\nOnce trust is lost, most users do not return.\n\nIf AI is meant to grow beyond a temporary trend and become a true industry,\n\nthe first thing to cultivate should not be performance, but trust.\n\nThis small change may seem minor, but it has the potential to create a large and lasting impact.",
  "title": "Feature Request: Add Explicit Reference Declaration for External Sources:"
}