What the Hate list means
The Hate card appears when at least one outbound link leads to a destination judged to be hate content. The list behind Review these links holds those links only. A link can have several categories, so the same link may appear under another card too.
There is no blocklist for hate. A link lands here in one of two ways:
- Word match. The destination’s title, description or page text contains “white supremacy” (or “white supremacist” and similar), “neo-nazi” or “neonazi”, or “hate group”. On its own this is graded Needs review.
- AI analysis. The destination is read together with the page that links to it and judged to be hate content. This can be graded Not safe. See verdicts explained.
The word list is short and describes groups by name. Hate material that uses other language is only found if the AI analysis catches it, so a clean Hate card is not proof that nothing is there.
When a hate link is legitimate
The phrases above are exactly what reporting and research use to describe hate, so many legitimate pages match:
- A news article or investigation about extremist groups.
- An anti-hate or anti-racism charity, or a monitoring organisation.
- A history site, museum or university course covering fascism and the Second World War.
- Academic research on radicalisation.
The report’s suggested action is to remove the links unless the hate content is the subject of the page. A history or campaigning page linking to those sources is on-topic, so no Out of context chip is added. On a site about anything else, it is.
When it is unwanted
A bakery, a sports club or a parish page that links to a hate group did not choose to. It may be injected code, a link left in a comment or a profile, a former contributor’s edit, or a domain that once hosted something harmless and was later taken over. Concealed links show the Hidden, Cloaked or Hidden in script chips; see hidden links.
Why it matters
- Reputation. A link can read as an endorsement. This is the category where that matters most to the people who visit.
- Visitors. Someone targeted by the group may be directly harmed by finding it linked from your site.
- Advertising. Ad networks and payment providers commonly restrict sites tied to hate content.
- Search and filters. Pages carrying unrelated links of this kind can be demoted, and schools, libraries and workplaces may block the site.
In your report
The card is titled Hate and carries a Medium risk chip. It is shown only when something was found. Its four figures are Flagged links, On a blocklist (always zero, as there is no list), Off-topic for their page and Hidden from visitors. The suggested action reads: remove the links to hate content unless it is the subject of the page.
Review these links opens the findings table filtered to this category, with columns Priority, Destination, Signals and Found on. Signals shows the chip hate, plus Out of context, concealment chips and Followed where they apply. Choose a row to see the anchor text and reason. See finding priorities for how findings are ranked.
Keyword match, needs a human look: hate group
/blog/parish-newsletter · Anchor text “local news”What to do
- Read the destinationOpen it in a private window. Decide whether it is reporting about hate or the hate itself.
- Find the link on your pageOpen the page in Found on and search for the anchor text, and for hidden or script-written copies.
- Remove it or keep it on purposeIf you did not choose the link, remove it. If it is a source for a page about the subject, keep it and mark the finding as reviewed.
- Find out how it got thereCheck recent edits, accounts, comments, profiles and plugins. If the site was compromised, follow After a WordPress hack.
- Check the restReview each row under the card and the other category cards.
Common mistakes
- Assuming the report has judged the destination’s views. It has matched words and context; a person decides.
- Marking a link nofollow and treating that as endorsing nothing. Visitors still follow it.
- Leaving old comments and profile links in place.
- Treating an empty card as a clear bill of health, given how short the word list is.
How to confirm it is fixed
Run a re-check from the report once the changes are live. It reads only the pages your flagged links are on, so it is quick, and it shows whether the hate links still appears. Marking a finding as reviewed is a note to yourself, not a fix: the next scan reads the page again and calls the finding resolved only when the link is gone or no longer flagged. If it is still there, it comes back as Still flagged. On a monitored site, later scans keep watching for it to return.
Questions
Why was an anti-hate charity flagged?
The word list matches the phrases that charities and journalists use. A word match is only Needs review, and the charity is on-topic for a page about the subject, so mark it as reviewed.
Is there a blocklist for hate sites?
No. Link Signals has lists for adult, gambling and violence only. Hate is found by reading the page.
Does Link Signals decide what counts as hate?
It matches a short set of phrases and asks an AI model to judge the page. A person has the final say on each finding.
Can I keep a link as a source?
Yes. Mark the finding as reviewed. The next scan reads the page again and shows whether the link is still flagged.
How this counts in your LinkSignal
A link to hate content is flagged as risky or unsafe, and that is what links that need review counts in your site’s external LinkSignal. If the same link is hidden, or sits on a page it does not fit, it is counted once in the hidden, cloaked and script-injected links or out of context links row instead. The LinkSignal overview lists every row, and how the score is worked out gives the points. LinkSignal is Link Signals’ own metric, worked out only from what this tool finds, and it is not connected to any search engine or third-party score.
See whether anything on your site links to hate content you did not choose.
