# AI Accountability

> A named person must answer for what an AI system does.

**Field:** Artificial Intelligence  
**Category:** Ethics & Regulation  
**Also known as:** Accountability (AI), Algorithmic Accountability, AI Responsibility

## Think of it like

It is like a shop sign with the owner's name. If something goes wrong, you know who to ask.

## Definition

AI accountability means a named person or group answers for what an AI system does. A model cannot be blamed or fined. So a human must own each system. That owner explains how it decides, fixes it when it goes wrong, and puts things right for people it harmed. It also means people can ask for a human to look again at a decision. Logs and records make this possible, because they show what happened.

## Example

A bank's AI turns down Lucy's loan. The bank has a team that owns the model. Lucy asks for a review, and a person checks her case and explains the result.

## Why it matters

With no owner, a bad mistake is nobody's job to fix. It just stays.

## Common mistake

Blaming the AI. "The computer said no" is not an answer, because someone chose to use it and must stand behind it.

## Three principles

- **A named owner.** One person or team for each system.
- **Explain and fix.** Say why, then put it right.
- **Keep records.** Logs show what happened.

## Related terms

- [AI Governance](https://learn.benibix.com/term/ai/ai-governance)
- [Responsible AI](https://learn.benibix.com/term/ai/responsible-ai)
- [Automated Decision-Making (ADM)](https://learn.benibix.com/term/ai/automated-decision-making)
- [Explainable AI (XAI)](https://learn.benibix.com/term/ai/explainable-ai)
- [Human-in-the-Loop](https://learn.benibix.com/term/ai/human-in-the-loop)
- [AI Regulation](https://learn.benibix.com/term/ai/ai-regulation)

![A decision about a person drawn as a chain: an AI model suggests yes or no, a highlighted human review can change it, the decision goes out, and it reaches the person it is about. A dashed arrow runs from that person back to the review, for asking a person to look again. Under it, a named owner answers for the whole chain.](https://learn.benibix.com/diagrams/ai4-who-answers.svg)

---

Source: https://learn.benibix.com/term/ai/ai-accountability
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-18
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
