Tools and testing
False positives and false assurance
- Last reviewed
- 9 August 2026
Guidance on how to check accessibility testing results and avoid false positives and false assurance.
[NEED TO ADD IMAGES AND CHECK WITH JAKE RE THE 30-40% STAT]
Understand false positives
A false positive is an issue flagged by an accessibility testing tool that is not actually an accessibility issue when you review it manually.
Accessibility testing tools cannot understand all the context of your content. For example, they may not be able to tell:
- whether the text in a link clearly describes its purpose
- whether an image needs alternative text
- whether an error message provides the right information
- whether a contrast issue applies to an exempt type of content, such as a logo
Contrast in images
Some accessibility testing tools may flag contrast issues for content that is exempt from the contrast requirements, such as logos and branding.
You should review the result manually to check whether the content needs to meet the relevant contrast requirements.
Understand false assurance
Accessibility testing tools cannot identify every accessibility issue. They can identify some issues automatically, but they cannot tell you whether your service works for everyone.
For example, a testing tool can check whether an image has alternative text. It cannot reliably tell whether the alternative text accurately describes the image.
An image with the following alternative text would pass an automated check:
alt="image123"
But this does not provide useful information to someone who cannot see the image.
Check alternative text
Alternative text should describe the purpose or meaning of an image where this information is important to understanding the content.
For example:
Do
A teacher standing in front of a class of children who are sitting at desks.
alt="A teacher standing in front of a class of children who are sitting at desks."
Don't
alt="image123"
or other filenames or descriptions that do not provide useful information.
There are exceptions where an image does not need alternative text, such as when it is purely decorative.
If you are unsure whether an image needs alternative text, use the [alt text decision tree] to help you decide.
Validate your test results
Do not rely on automated testing alone. Review flagged issues manually and carry out additional manual and specialist testing where appropriate.
Testing with people who have different access needs can also help you identify barriers that automated tools cannot detect.
See [Find and use tools] for guidance on the tools and techniques you can use to test for accessibility.