Check whether someone’s claimed qualifications are real with ChatGPT

← All use cases / Learning

ToolChatGPT

Wondering whether the author of a website really had the PhD and teaching experience they claimed, a Hacker News commenter asked ChatGPT to look for evidence from the institutions named. In about 30 seconds it pointed them to university pages backing up the claims, including a list of past PhD recipients and a university page linking to the author's personal site.

Prompt to copy
The author of [WEBSITE, ARTICLE OR PROFILE LINK] says they have [CLAIMED QUALIFICATION OR ROLE] from [INSTITUTION OR ORGANISATION]. Please look for evidence from independent, official sources, such as the institution's own website, a public register or a professional body, that supports or contradicts this. List each source with a link and one line on what it shows. Tell me if you couldn't find anything, and don't treat the person's own website as proof on its own.

Replace the [PLACEHOLDERS] with your own details before you send it.

How to try it

  1. Note exactly what the person claims, including the qualification, the institution and any dates.
  2. Ask ChatGPT to search for evidence from official sources, using the prompt above.
  3. Open every link it gives you and read the relevant page yourself.
  4. Look for two-way links: the institution pointing to the person, not just the person pointing to the institution.
  5. If you still can’t confirm it, contact the institution or professional register directly.

Why this works

Anyone can write ‘PhD’ or ‘qualified expert’ on a website, and it’s sensible to check before you trust their advice, hire them or share their work. Checking used to mean trawling through university pages and registers by hand. A search-enabled chat can do that legwork quickly and hand you the links, and as the commenter found, it can be done in under a minute. The key is that you still open the sources and judge them yourself. Our guide to spotting fake images and news with AI covers more ways to check what you read online.

Tips to get a better result

  • Ask for official registers in your country for regulated jobs, such as builders, accountants or therapists.
  • Check the dates match: a degree awarded after someone says they started teaching is worth a closer look.
  • If someone is pushing you to pay quickly, our guide to spotting scams with AI can help.
  • Save the links you find, in case you need to refer back to them.

Common mistakes

  • Trusting the AI’s summary without opening the sources; it can mix up people with similar names.
  • Treating a lack of results as proof of a lie; some records simply aren’t online.
  • Sharing accusations publicly before you’ve confirmed the facts.

Try a variation

Here’s a variation of our own to try next:

I'm about to hire [TRADE OR PROFESSIONAL] called [NAME OR COMPANY] in [AREA]. Tell me which official registers or accreditation schemes I should check for this kind of work, with links, and what to look for on each.

Related

More learning use cases

Turn a vague guitar goal into daily practice reminders with ChatGPT

A TechRadar writer who had owned a guitar for years but never stuck to practising asked ChatGPT's Tasks feature for a plan to get good over the next year. It set up recurring daily reminders split into different kinds of practice, added ideas like finger-strengthening exercises and recording yourself to hear your progress, and suggested new songs to match their skill level.

Full guide →
Source: TechRadar

Learn to recognise local plants and wildlife with a ChatGPT-made Anki deck

One person who wanted to get better at naming the plants and animals around them asked ChatGPT to build an Anki deck. In one go it gathered 160 common local plants, fish and animals, wrote short descriptions, added photos and packaged it as a single download.

Full guide →

Learn Excel with a four-week plan built around your own reports with Claude

Someone asked Claude to teach them Excel the way an expensive private tutor would. Rather than a generic course, it built a four-week learning plan around the reports they already make at work.

Full guide →

Reference: Hacker News (ndriscoll)