Hallucinations & Over-reliance: The AI Danger Explained

What is the danger?

AI models regularly invent facts, sources and quotes, and present them with total confidence. As people rely on AI for legal, medical, financial and educational answers, these errors cause real harm — and constant reliance may weaken our own thinking skills.

How it happens

Models predict plausible text, not verified truth.
Fake references and citations that look completely real.
People stop checking because the AI “is usually right”.

Real cases

Lawyers in the US case Mata v. Avianca were sanctioned in 2023 after submitting a court filing with fake case citations invented by ChatGPT.
Source: Reuters ↗
A Canadian tribunal ordered Air Canada to honour a refund policy its website chatbot had invented (2024).
Source: BBC ↗

Warning signs

Very specific facts or quotes with no link you can open.
Answers that sound right but you cannot verify anywhere else.

What you can do

Always verify important answers — especially medical, legal and financial ones.
Ask the AI for sources, then open them yourself.
Keep doing hard thinking yourself; use AI as a helper, not a replacement.

Frequently asked questions

What is the danger of hallucinations in AI?

How does it happen?

Models predict plausible text, not verified truth. Fake references and citations that look completely real. People stop checking because the AI “is usually right”.

How can I protect myself?

Always verify important answers — especially medical, legal and financial ones. Ask the AI for sources, then open them yourself. Keep doing hard thinking yourself; use AI as a helper, not a replacement.

Hallucinations & Over-reliance