Hate Speech

Hate speech is content that attacks, dehumanizes, or demeans people based on protected characteristics such as race, ethnicity, religion, gender, sexual orientation, or disability. It targets people for who they are rather than anything they have done.

What it means

What separates hate speech from ordinary insults is the basis of the attack. An insult targets an individual over their behavior or an argument. Hate speech targets a person or group because of an identity trait, which is why nearly every platform and many legal systems treat it as a distinct and serious category. It ranges from explicit slurs to coded language and dog whistles that mean the same thing without saying it.

Real-world examples

• A slur directed at someone because of their ethnicity or religion.
• "People like you do not belong in this country," aimed at a group's origin.
• A number code or symbol that a community uses as a stand-in for a slur, which only implicit moderation will catch.

Why it matters

Hate speech drives targeted people out of a community and exposes platforms to legal and reputational risk, especially in regions with strict laws. It is also a favorite target for evasion, since users who spread it quickly learn to disguise it. Effective detection has to understand meaning and context, not just scan for a list of forbidden words, or it will miss the coded version and punish clinical or quoted uses. Supervisor labels this category as Hate or Racism.