British Sign Language Glossaries of Curriculum Terms

BSL AI Glossary - Explainability definition

Translation: Explainability in AI means being able to understand WHY an AI system made a particular decision or gave a particular answer.

Imagine you apply for a job, and an AI system says NO. Explainability means you can ask — WHY did it say no? What did it look at? What made it decide that? Without explainability, you just get the answer with no reason. That's not fair.

A lot of AI works like a BLACK BOX. Information goes IN, a decision comes OUT — but nobody can see what happened in the middle. Not even the people who built it sometimes. A Black Box AI will give you an answer but no explanation and Explainable AI will give you an answer AND tells you why. Why is it important?

• Healthcare: If AI looks at a medical scan and says someone is ill — doctors need to know WHY it thinks that. They can't just trust it without a reason.
• Justice: In some countries AI is used to decide if someone should get bail or go to prison. If it says no — that person deserves to know why. A machine making life changing decisions with no explanation is very dangerous.
• Education: If AI marks your work and gives you a low grade — you should be able to find out what it thought was wrong. How else can you improve?
• Jobs: Many companies now use AI to sort job applications. If AI rejects you — was it fair? Was it biased? Without explainability you will never know.

Explainability in AI means you have the right to understand why an AI made a decision that affects you. A fair AI should never just say YES or NO without being able to show its working.