Hate/Discrimination: the web filtering category for content that targets people for who they are
This category labels sites whose main purpose is to promote hatred or discrimination against people because of a protected characteristic. It is part of the CIPA-relevant set in our database of 59 categories.
It is also one of the categories where getting the border right matters most. Reporting on hate, research and anti-hate work must stay reachable.
What the Hate/Discrimination category includes
A site carries this label when attacking or demeaning a group of people is what the site is for, or a substantial part of what it publishes. One offensive comment on an otherwise ordinary site does not qualify.
Supremacist advocacy
Sites arguing that one group is superior and others are lesser, or should lose rights.
Dehumanising propaganda
Material that portrays a group as vermin, disease or a threat to be removed.
Recruitment for hate movements
Pages that invite readers to join or support organised groups built around hatred of others.
Hate merchandise
Shops whose main stock is hate symbols, slogans or propaganda items.
Hate-focused communities
Boards and groups where demeaning a target group is the shared purpose.
Atrocity denial as propaganda
Sites denying or praising genocide or mass atrocities to promote hatred of the victims' group.
Targeted harassment campaigns
Pages organising abuse against people because they belong to a protected group.
Hate media distribution
Sites distributing music, videos or memes produced to spread hatred of a group.
What it excludes, and why news and anti-hate organisations stay out
The same words and images can appear on a hate site and on a site that exposes it. The label follows what the site does with them, not which topic it covers.
Labelled Hate/Discrimination
- A site that promotes a supremacist ideology
- A board where members trade slurs about a religious group
- A shop selling hate symbols as its main product line
- A page denying an atrocity to stir hatred of its victims
- A campaign page organising abuse of a minority group
Not labelled Hate/Discrimination
- News reporting on hate crimes or hate groups: News/Media
- Civil rights and anti-hate organisations
- History education about persecution and genocide: Education/Reference
- Academic research on extremism and prejudice
- Anti-discrimination law and government guidance: Government & Legal
Why these exclusions matter
Students study hate to understand it
History, civics and literature lessons cover persecution and propaganda. Blocking the sources that explain them defeats the lesson.
Targets of hate need help
People facing abuse look for reporting lines, advice and support groups. Those sites must stay reachable on every network.
Journalism exposes hate
Reporting on hate groups is how the public learns about them. A filter that blocks it hides the problem rather than reducing it.
Professionals counter hate
Trust and safety teams, educators, researchers and police need the material that explains how hate spreads.
Political debate
Arguments about immigration, religion in public life or equality law go to Political, unless they attack people for who they are.
Religious teaching
Statements of belief go to Religion & Spirituality. The label is added only when a site campaigns against people.
Violent extremism
Content calling for attacks also carries Violence, so either rule can act on it.
General rudeness
Insults, profanity or offensive humour without a protected-group target are not hate in this sense.
How a site is judged: the decision path
Sites are categorized from their content, reviewed and validated. The questions below show the reasoning behind the label in plain terms.
- Is a group of people being targeted?The content must be aimed at people defined by a protected characteristic, not at an idea, a policy or a government.
- Does the site promote, or report?Promotion means the site endorses the hostility. Reporting, analysis and quotation for criticism point to another category.
- Is it a main purpose of the site?The content has to describe a real share of the site. Isolated user posts on a large platform do not make the whole domain hateful.
- Is violence called for or celebrated?If yes, Violence is added as well, ranked by how much of the site it describes.
- What else is on the site?Other labels, such as Community Sites, Forums / Message Boards or Shopping, are kept so the full picture reaches the policy engine.
Grey areas and the usual outcome
| Situation | Usual label | Reasoning |
|---|---|---|
| Archive of historical propaganda with scholarly notes | Education/Reference | Context and commentary show the purpose is study. |
| Same propaganda reposted without comment to recruit | Hate/Discrimination | The purpose has changed from study to promotion. |
| Comedy site with occasional offensive jokes | Arts & Entertainment | Not a site built to promote hatred of a group. |
| Comedy site devoted to mocking one ethnic group | Hate/Discrimination | Demeaning a protected group is its main output. |
| Campaign against a law or government | Political | Criticism of institutions is not hate against people. |
| Site tracking hate incidents for victims | Community Sites or Government & Legal | Supporting targets is the opposite of promoting hate. |
Accuracy: Hate/Discrimination is part of the CIPA-relevant set. Regular manual validations measure accuracy, and validation reports for these categories are available to licence holders, quarterly or monthly on request.
Typical site types
The category is small compared with mainstream categories, and many of its domains are short-lived. Common shapes are listed without naming any site. We never link to them.
| Site type | What it looks like | Usual extra labels |
|---|---|---|
| Ideology sites | Long articles and "reading lists" presenting a supremacist world view | Political, sometimes Religion & Spirituality |
| Hate boards | Discussion boards organised around contempt for a group | Forums / Message Boards |
| Propaganda blogs | Single authors publishing hostile posts and memes | Blogs/Personal Sites |
| Merchandise shops | Stores whose main stock is hate symbols and slogans | Shopping |
| Media archives | Collections of hateful music, video or images | Streaming Media, File Sharing/P2P |
| Movement pages | Recruitment and event pages for organised hate groups | Community Sites, Violence |
| Mirror and backup domains | Copies of removed sites on new domains | Same labels as the original |
Short-lived domains: hateful content often moves to new domains after takedowns. About 300,000 newly registered domains are categorized every day and checked again after a week and a month, so mirrors are picked up as their content appears.
Who blocks or allows Hate/Discrimination: policy table
This category is blocked by default almost everywhere. The real policy work is in deciding who needs an exception and in keeping the excluded categories open.
| Environment | Default | Reasoning | Exceptions to plan |
|---|---|---|---|
| K-12 schools (CIPA) | Block | CIPA's filtering duty covers obscene images, child pornography and images harmful to minors. Hate is not named, but schools block it under their internet safety policy. | Keep News/Media and education sources about hate open for lessons |
| Public libraries | Block for minors | Libraries weigh patron access to lawful material. CIPA lets libraries disable filtering for adults for bona fide research or other lawful purposes. | Adult unblock requests; check local policy and current legal advice |
| Businesses | Block | Hostile content at work creates harassment and conduct risks. | Trust and safety, HR investigations, security research |
| Media and research | Allow by group | Journalists and researchers need to see the material they report on. | Named groups with logging. See media web filtering |
| Law enforcement | Allow by group | Investigators monitor hate movements as part of their work. | Isolated browsing for investigators. See law enforcement web filtering |
| Guest Wi-Fi | Block | Venues do not want hateful content on their network or screens. | Rarely needed |
| Parental control | Block, all ages | Part of every child preset in most apps. | Parents may allow specific education sources |
| ISP family profiles | Block in family profile | A standard part of opt-in family filtering. | Off in unfiltered profiles |
For school category sets and CIPA background, see CIPA web filtering and K-12 school web filtering. For libraries, see library web filtering. Schools outside the US follow their own national guidance, so check the current requirements.
Multi-category cases
Hate content rarely lives alone. Ranked multiple labels let a policy act on the part that matters and leave the rest of a mixed site visible in reports.
Hate movement with calls to attack
Blocked by any profile that blocks either label.
Board built around contempt for a group
The forum label helps reports; the hate label drives the block.
Hate symbol shop
Ranked hate first because the stock defines the site.
Group that started as a hobby club
When the main output turns hostile, the hate label is added. See Community Sites.
Party site using hateful rhetoric
Ranking reflects how much of the site the hostile content takes up.
Museum exhibit on propaganda
No hate label: the site explains propaganda, it does not spread it.
Investigative newsroom series
Reporting stays news, even when it quotes hateful material.
Cult-like group attacking outsiders
Both labels let schools and family apps block it while faith sites stay open.
How vendors use the Hate/Discrimination category
Products that license the database treat this as a liability category with a few role-based exceptions.
School filters
On by default in every student profile, with reporting so staff can see repeated attempts and respond with care.
Parental control apps
Part of every child preset. Some apps alert parents to repeated visits rather than only blocking.
Secure web gateways
Blocked company-wide, with exempt groups for trust and safety or investigations, often through isolated browsing.
DNS filters
Included in family and school resolver profiles. See DNS filtering vs web filtering.
Brand safety tools
Ad platforms use the label to keep ads away from hateful pages.
ISP filtering
Part of opt-in family profiles. See ISP web filtering.
Delivery: CSV, Parquet or JSON files that run in your own systems, daily deltas for new domains and a lookup API with licences. Check a single domain with the domain lookup tool.
Setting this category up well: admin checklist
Blocking hate is simple. Blocking it without blocking the material people need to fight it takes a little more care.
Do
- Block Hate/Discrimination and Violence together in student and child profiles
- Keep News/Media, Education/Reference and Government & Legal open
- Decide whether your rule reads the first label only or any label
- Create named exception groups for staff who investigate or research hate
- Use isolated browsing for those groups where possible
- Review blocked-request logs with a pastoral or HR lead, not only IT
Avoid
- Keyword blocks on words like a group's name, which catch support and news sites
- Blocking civil rights and anti-discrimination organisations by mistake
- Treating all political or religious sites as hate
- Giving whole departments an exception when a few people need it
- Copying a student profile to adult library terminals without review
- Ignoring repeat attempts that may signal a welfare concern
When a student keeps trying to reach hate sites
Report the pattern
Filter reports can show repeated blocked attempts. One attempt is often curiosity; a pattern may need a conversation.
Check the source
Look at which sites were requested and whether a lesson or assignment explains them before drawing conclusions.
Involve the right staff
Safeguarding or pastoral staff are better placed than IT to talk with a student about what they found.
Keep proportionate notes
Follow your school's data protection rules when storing browsing records about named students.
Hate/Discrimination FAQ
What does the Hate/Discrimination category cover?
Sites whose main purpose is to promote hatred, contempt or discrimination against people because of a protected characteristic such as race, religion, disability, sex, gender identity or sexual orientation.
Why are news sites and anti-hate organisations not in this category?
They report on, study or counter hate rather than promote it. Blocking them would remove what students, researchers and staff need to understand and respond to hate.
Does CIPA require schools to block hate sites?
CIPA's filter requirement covers obscene images, child pornography and images harmful to minors. Hate is not named, but most schools block this category under their internet safety policy.
Is political debate about immigration or religion classified as hate?
Not by itself. Debate about policy, laws or beliefs goes to Political or Religion & Spirituality. The label applies when content attacks people for who they are.
How does this category relate to Violence?
Content that calls for or celebrates violence against a group carries both labels, so a policy that blocks either acts on it.
Does one hateful user post make a whole platform hateful?
No. The label describes the site's main purpose and a real share of its content. Large platforms are judged on the site as a whole.
Is accuracy measured for this category?
Yes. It is part of the CIPA-relevant set, and validation reports with measured accuracy are available to licence holders.
Can a site lose the Hate/Discrimination label?
Yes. Domains are refreshed monthly. If a domain changes hands or its content changes, the labels change with it, which matters for expired domains that get new owners.
Why not block by keywords instead?
Keyword rules catch news, history and support sites that use the same words. Category labels follow the purpose of the site, which keeps those sources open.
Where can I see the other categories?
The full web filtering categories list covers all 59.
License a carefully bounded hate category
Hate/Discrimination with ranked multi-category labels across 120 million domains, validated as part of the CIPA-relevant set. Starter database tiers from $999; enterprise and OEM licences are quoted per deployment.