---
title: "Opacity"
description: "The inability of a person to follow how a system reached an output."
canonical: https://aigovernanceengineer.com/glossary/opacity
author: "Jorge García Aibar"
license: "CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)"
doi: https://doi.org/10.5281/zenodo.22956197
version: "0.5.0"
updated: 2026-09-24
---

# Opacity

The inability of a person to follow how a system reached an output. It has three sources (secrecy, technical illiteracy, and the nature and scale of machine learning) [1], each with a different fix: disclosure, literacy, and explanation methods plus behavioural evals.

- Developed in: [ch. 11, Eight characteristics that break classic IT governance](https://aigovernanceengineer.com/bok/ai-defined#eight-characteristics-that-break-classic-it-governance); [ch. 11, Contrast pairs](https://aigovernanceengineer.com/bok/ai-defined#contrast-pairs)
- Chapters: [ch. 11, AI Defined](https://aigovernanceengineer.com/bok/ai-defined)
- Contrast with: [Explainability](https://aigovernanceengineer.com/glossary/explainability)
- In the glossary chapter: https://aigovernanceengineer.com/bok/glossary#t-opacity

## Sources

[1] "How the machine thinks: Understanding opacity in machine learning algorithms" (three forms of opacity; Big Data & Society 3(1)). SAGE (Jenna Burrell). 2016-01-06. https://doi.org/10.1177/2053951715622512 (verified: primary)