Skip to content

AI Safety & Responsible AI

Counterfactual explanation

An explanation of what would need to change for an AI to reach a different decision.

Example

A tool shows which modeled factors would change a rejected application to an accepted one.

Why people use it

It helps people see which changes would alter an AI's decision.

What you'll hear

“What would need to change for a different answer?”

What this means for you

Ask whether the proposed changes are realistic and within your control.

Can you control it?

Sometimes

Sometimes. Your choices depend on the tool and your access. The settings available to an everyday user may differ from those available to the people running it.

Common questions

Could the suggested change be impossible for someone to make?
Yes. A system may suggest changing something the person cannot control, such as their age or past history.
Can there be more than one explanation?
Yes. Different combinations of changes may produce the same different decision.
Can it suggest something a person cannot change?
Yes. An explanation may mention age or past events unless the system rules those changes out.

Related terms

Still have questions?

Up to 500 characters.

Ask LATHIC about AI. Relevant glossary entries may be included.

Your question, the glossary entries it matches, and a rotating pseudonymous identifier go to Microsoft Azure’s OpenAI service through Vercel AI Gateway to generate an answer. Zero retention and no training are required of the provider, and LATHIC does not save your question or answer. Privacy Notice