What Is Human-Centred AI?

Human-centred AI is an approach to designing and using artificial intelligence that starts with the needs of the people who will actually interact with it.

That might sound obvious, but it changes the questions organisations ask when introducing AI. Instead of starting with “What can we automate?”, a human-centred approach might begin by looking at the problem people are trying to solve, where AI could genuinely help and where human involvement still matters.

For example, an AI system used in customer service might handle repetitive administrative work while leaving more complicated or sensitive conversations to people. In healthcare, AI might help a clinician identify patterns in medical information without making the final decision about a patient’s care.

Human-centred AI can also mean thinking about the experience of the person using or affected by the system;

• Is it easy to understand?
• Can someone question or challenge its output?
• Does it make their job easier, or has it simply introduced another system they now have to manage?

There isn’t one single feature that makes an AI system human-centred. It’s more about the way the technology is designed and introduced, with attention paid to the people around it rather than treating the technology itself as the starting point.