---
title: 'False negative'
description: 'An outcome incorrectly classified as absent when it is actually present.'
canonical_url: 'https://darkfactory.dev/glossary/false-negative'
markdown_url: 'https://darkfactory.dev/glossary/false-negative.md'
collection: glossary
date_published: '2026-08-03T00:00:00-04:00'
date_modified: '2026-08-26T00:00:00-04:00'
---

# False negative


## Definition

A false negative occurs when a system predicts that a condition is absent even though the reference says it is present. It is a miss: undetected fraud, a disease not flagged, a relevant document not retrieved, or a failing test judged as passing.

The false-negative rate is `false negatives / (true positives + false negatives)`. It answers: among the actual positive cases, what share did the system miss? The denominator requires trustworthy reference labels.

The operational harm depends on recovery. A missed spam message may be noticed by the user; a missed safety defect may become irreversible. Evaluate false negatives by relevant slice and consequence, not only as an aggregate count.

## Distinguish it from nearby terms

A false positive raises an alarm for a condition that is absent. A false negative misses one that is present. Recall is one minus the false-negative rate when both use the same positive class and reference set.

## Check your understanding

A content filter misses 2 percent of prohibited items overall but 20 percent in one language. The aggregate false-negative rate hides the population carrying most of the risk.

## Related terms

- [False positive](https://darkfactory.dev/glossary/false-positive)
- [Recall](https://darkfactory.dev/glossary/recall)

## Evidence and further reading

- [NIST AI Resource Center Glossary](https://airc.nist.gov/glossary/)
- [Google Machine Learning Glossary](https://developers.google.com/machine-learning/glossary/)