What is “AI accountability”

A simple explanation of AI accountability and why it matters

AI accountability is about who is responsible for the outcomes produced by AI systems. When a system makes a decision or generates an output, it’s not always obvious where that responsibility sits.

Is it the developer who built the system, the organisation that deployed it, or the person using it?

The question becomes more pressing as AI is used in situations that directly affect people, such as hiring, healthcare, or financial decisions. When something goes wrong, it matters who can explain what happened, justify the outcome, or put it right.

In practice, responsibility is rarely held by just one party. It tends to sit across multiple layers.

So the question becomes: when decisions are distributed across people and systems, who is ultimately accountable?