False Positives in Audit Tools: Reports Checked Against the Rules
We run other companies' audit tools on our own site, aiwebpageseo.com, then check every finding they show against two things: the code on the live page, and the published rule the finding is based on. When the page meets the rule the tool says it breaks, that finding is a false positive.
Each review below links to the full write-up, with the code, the rule and who complies.
The reports we have checked
| Tool | Date | What it reported | What checking found |
|---|---|---|---|
| accessScan (accessiBe) | 13 Oct 2026 | "Accessible", then 5 failures and 2 best-practice warnings | All 5 failures and 1 of the 2 warnings were false positives: content it said was outside the main landmark is inside it, decorative icons correctly hidden, links already named. The other warning is optional advice. |
| Recite Me free scan | 12 Oct 2026 | 186 issues: 105 at Level A and AA, 81 at AAA | Every Level A and AA example shown was a false positive. The 81 AAA contrast items were accurate, but WCAG does not recommend requiring AAA. |
| seoClarity AI Search Readiness Checker | 12 Oct 2026 | Score 98/100, "Low Risk" | Took 2 points off for 298 characters added by JavaScript: a 250-character cookie banner, which UK rules require, and two live status lines. All the content was in the HTML. |
| aisearchradar.io | 13 Oct 2026 | Score 81/100, 12 critical items | Recommended adding ai.txt, Person markup, sameAs and a last-updated date, all already on the site; promised HowTo results Google stopped showing in 2023. |
| Roar Digital Search Radar | 13 Oct 2026 | Score 42/100, "Weak visibility" | Placed the business in Sheffield instead of Chesterfield, tested local city searches for an online service and marked consent-first analytics as missing. |
How we check a report
- Keep the original. A dated copy of every report, as the tool showed it.
- Check the code. Each finding is compared with the live page's code, quoted in the write-up.
- Check the rule. Each finding is judged against the rule it rests on, linked so anyone can read it.
- Say who complies. Every write-up states whether the page meets the rule and whether the tool's check matches it.
- Keep opinion separate. Anything that is our view rather than a checked fact is labelled as our view.
The rules we check against
- W3C: Web Content Accessibility Guidelines (WCAG) 2.2
- W3C: Techniques for WCAG 2.2
- W3C: WAI-ARIA 1.2
- Google Search Central: Organization structured data
- schema.org
- ICO: Guidance on the use of storage and access technologies
- GOV.UK: Public sector website and mobile application accessibility monitoring
Frequently asked questions
What is a false positive in an audit report?
A false positive is a problem an audit tool reports that is not actually there. For example, a scanner may say an icon has no text alternative when the icon is decorative and correctly hidden from screen readers, which is exactly what WCAG asks for. Acting on false positives wastes time and can make a page worse, so each finding should be checked against the page code and the rule it cites.
How do you check another tool's report?
We run the tool on our own site, keep a dated copy of the report, and check every finding it shows against two things: the code on the live page, and the published rule the finding is based on, such as a WCAG success criterion, a W3C technique or Google's structured data documentation. A finding counts as a false positive only when the code meets the rule the tool says it breaks.
Why do free scanners report so many false positives?
Many free scanners read the HTML source and match patterns, so they flag an image without alt text even when it is hidden, or any tabindex attribute whatever its value. They do not see the page as a browser or screen reader does. Some also count optional Level AAA items, or best-practice advice, as failures, which makes a headline total look far worse than the page is.
Does a report with no false positives mean a site is compliant?
No. Automated tools, ours included, can test many WCAG criteria but not all of them: 20 of the 55 Level A and AA criteria in WCAG 2.2 need a person to judge, such as whether alt text describes each image well. A clean automated report is the right starting point for a manual review, not proof of compliance.
Check your own site against the rules
Our Accessibility Audit checks a page against all 55 WCAG 2.2 Level A and AA criteria as visitors see it, for £2.49, and lists best-practice advice separately so it never inflates the count.