Bias And Fairness
AI can pick up unfair patterns from its training data; learn to spot bias and use AI more fairly.
Where bias comes from
AI models learn from huge piles of text and images created by people. If that material reflects unfair patterns from the real world, the model can repeat them. For example, an AI asked to picture a 'nurse' or a 'CEO' may lean on old stereotypes about who does those jobs, simply because that is what showed up most in its training data.
How bias shows up in answers
Bias is not always obvious. It can appear as an AI that writes warmer feedback for some names than others, suggests harsher tones toward certain groups, or assumes a person's role based on gender. Because the wording sounds confident and fluent, unfair assumptions can slip by unless you are watching for them.
Using AI more fairly
You can reduce harm with a few habits. Read AI output with a critical eye, especially for anything that affects people, like hiring notes or reviews. Ask the AI to consider multiple perspectives, or to avoid assumptions about gender, age, or background. For high-stakes decisions about people, keep a human in charge and never let the AI decide alone.
Key takeaways
- Bias comes from unfair patterns in human-made training data.
- Fluent, confident wording can hide unfair assumptions.
- Read critically and keep a human in charge of decisions about people.
4 questions · pass at 60% to earn XP