About this Handout
A Beginner's Guide to Understanding AI and Cognitive Biases
Generative AI chatbots are fast, fluent, and available around the clock, qualities that make them useful and also make it easy to over-trust what they say. This guide walks through 14 cognitive biases most likely to distort judgment in everyday AI chat use, from automation bias and overreliance to sycophancy, anchoring, anthropomorphism, and social-proof effects.
Each bias is presented on its own page with a plain-language explanation and a real, sourced example, drawn from court filings, peer-reviewed studies, and reported incidents, so readers can see how the bias actually plays out rather than just how it's defined. The biases are ordered as a qualitative risk ranking (estimated frequency times potential severity), with a note that actual risk shifts by user, system design, and subject matter, especially in healthcare, finance, employment, law, safety, and personal relationships.
The guide is intended as an accessible entry point for teams and individuals who use AI chat tools regularly but haven't had a framework for spotting where those tools can quietly mislead. It reflects En Dash Consulting's broader approach to holistic, ethical, human-centered AI governance: helping organizations adopt AI thoughtfully, with clear eyes about its limitations.