Tip: press / anywhere to search. Or open a random note →

Handwritten notebook page explaining why AI makes things up, known as hallucinations, with tips to protect yourself
Tap the note to see it full size

AI & Tech One-page note

Why AI Makes Things Up ("Hallucinations")

AI can make things up and sound 100% sure. Here's why, from the people who build it. Save this page.

What's on the page

The note as text, so you can read it, copy it or use a screen reader.

What is a hallucination?

  • When AI gives an answer that sounds right but is false
  • Fake facts, fake quotes, fake sources

Why it happens (OpenAI research, 2025)

  • AI learns by predicting the next word
  • Common patterns (spelling, grammar) are easy to learn
  • Rare facts (like someone's birthday) can't be guessed from patterns

Why it keeps happening

  • AI is often tested like a quiz
  • Guessing can score points. "I don't know" scores zero
  • So models learn to guess

Protect yourself

  • Ask for sources, then open them
  • Double-check numbers, names, dates and quotes
  • Be extra careful with health, money and legal questions
  • Tell it: "Say if you're not sure"

Mistakes

  • Trusting it because it sounds confident
  • Copying AI citations without checking they exist
Want the full story? AI Hallucinations: Why AI Makes Things Up | Scribbled Smart Read the full guide →

Sources

  1. OpenAI: Why language models hallucinate
  2. arXiv 2509.04664: Why Language Models Hallucinate (Kalai, Nachum, Vempala and Zhang, 2025)
  3. Judiciary of England and Wales: Ayinde v London Borough of Haringey, and Al-Haroun v Qatar National Bank (High Court, 6 June 2025)

How we checked this

Every fact on this note comes from the 3 sources listed here, checked against official or peer-reviewed sources where possible. AI tools help with drafting; a person checks every fact against the sources. How we research

Spotted a mistake? Tell us and we'll fix it and log it on the corrections page.

More from AI & Tech

All AI & Tech notes →