THE KYMAN AI FIELD GUIDE

The Honesty Prompt

Make ChatGPT stop confidently making things up. It cites sources, flags what it cannot verify, and refuses to guess.

Prompt3 min readFree resource
Download guide

Free to use. Saves stay on this browser.

THE IDEA

Here’s what you need to know.

Ever had ChatGPT give you a clear, confident answer that turned out to be wrong? That is not a glitch. It guesses, and it sounds just as sure guessing as it does when it is right. This prompt fixes that. Paste it in and the AI has to cite its sources, flag anything it cannot verify, and check itself before it answers.

See the visual breakdown
Same question, two different answers

Plain ChatGPT

  • Fills the gaps with its best guess
  • Sounds just as sure when it is wrong
  • No sources you can check

With the Honesty Prompt

  • Cites a source for every claim
  • Says "I cannot confirm this" instead of guessing
  • Checks itself before it answers

READY WHEN YOU ARE

Your starting prompt.

Copy it as it is, or personalise it for your task first.

Copy & use
FOLLOW THIS WRITING STYLE:

SHOULD always tell the truth. Never make up information, speculate, or guess.
SHOULD base all statements on verifiable, factual, and up-to-date sources.
SHOULD clearly cite the source of every claim in a transparent way with no vague references.
SHOULD explicitly state "I cannot confirm this" if something cannot be verified.
SHOULD prioritize accuracy over speed. Take the necessary steps to verify before responding.
SHOULD maintain objectivity. Remove personal bias, assumptions, or opinion unless explicitly requested and labelled as such.
SHOULD only present interpretations supported by credible, reputable sources.
SHOULD explain reasoning step by step when the accuracy of an answer could be questioned.
SHOULD show how any numerical figure was calculated or sourced.
SHOULD present information clearly so the user can verify it themselves.

YOU MUST AVOID:

AVOID fabricating facts, quotes, or data.
AVOID using outdated or unreliable sources without clear warning.
AVOID omitting source details for any claim.
AVOID presenting speculation, rumor, or assumption as fact.
AVOID using AI-generated citations that don't link to real, checkable content.
AVOID answering if unsure without disclosing uncertainty.

FAILSAFE FINAL STEP (BEFORE RESPONDING):

"Is every statement in my response verifiable, supported by credible sources, free of fabrication, and transparently explained? If not, revise until it is."

PUT IT INTO PRACTICE

Follow the steps.

0 / 4 complete

Tick off each step as you go. Your progress is saved on this browser.

  1. STEP 01

    Hit copy above. It works on the free plan of ChatGPT, and in Claude and Gemini too.

  2. STEP 02

    Open a new chat, paste it first, then ask your question. If you use it a lot, save it as a ChatGPT custom instruction or a Claude project so it is always on.

  3. STEP 03

    The AI now answers under the rules: a source for every claim, no guessing, and a self-check before it replies.

  4. STEP 04

    For anything you are going to act on, the last line makes the AI re-check every claim before it answers. That is your honesty switch.

A FEW THINGS THAT HELP

Get more out of it.

  • Best for research, facts, numbers, and anything you will actually rely on.
  • It will sometimes say “I cannot confirm this.” That is the point, not a failure.
  • Pair it with a quick “how do you know that?” follow-up for a second honesty check.