AI engineering glossary · AI ethics law product
What is accountability (AI)?
Accountability (AI) (also called algorithmic accountability): The principle that identifiable people or organisations must answer for what an AI system does, with clear ownership, records and ways to correct harm. It rules out blaming the algorithm and requires named responsibility for outcomes.
Accountability (AI): an example
A company assigns a named product owner responsible for its loan model, with logs that show who approved changes and why.
Related terms
More AI ethics law product terms
- Activation rate
- AI audit
- AI governance
- AI incident
- AI Incident Database
- AI risk assessment
- Algorithmic bias
- Algorithmic impact assessment
Where you learn it
Accountability (AI) is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.