13 AI failure patterns and prompts to catch them
AI can mislead people in more ways than simply making up facts. It may agree with a bad idea, invent that were never checked, claim a task is finished when it is still incomplete, or apologize and then repeat the same mistake.
The material collects 13 types of AI “lies” and pairs each one with a prompt meant to catch that behavior. The list is open to additions if readers have seen another type of misleading AI response.
Key points
- AI may agree with flawed ideas instead of challenging them.
- AI may invent that were not actually verified.
- AI may say work is complete when only part of it is done.
- AI may apologize and then repeat the same mistake.
- The list pairs each failure with a prompt designed to catch it.