EASI For Non-ProfitsEnabling Augmented Social Impact

Episode 2 of 5

✎ episode 2

AI Lies (Confidently)

AI never says 'I don't know.' It says wrong things in a confident voice. Here's how to catch it.

Watch 6–8 min

Key takeaways

  • AI generates plausible text, not verified facts
  • Confidence is not accuracy
  • Verify anything with names, numbers, dates, laws
  • Ask "what's your source?" and check it yourself

Try it now · 2 minutes

Ask AI a factual question about your own organisation. Spot what it got wrong or invented.

Check yourself

This episode's 3 quiz questions cover:

  • Spotting confident but wrong answers (hallucination)
  • Verifying names, numbers and dates yourself
  • Asking AI for its source and checking it

You'll answer these in the final quiz.