The Instagram Algorithm Encyclopedia

Topics · Policy

Hate speech

Contents
  1. What Instagram prohibits and how violations are enforced
  2. How hate speech is detected and measured
  3. How it has changed
  4. Controls for limiting exposure to abuse
  5. How hate speech affects recommendations and features
  6. See also
  7. References

Hate speech on Instagram is an attack on people based on protected characteristics, such as race or religion, and is prohibited on the platform1. Meta responds to reports and uses technology outside private messages to look for content that might violate its rules2. People can also use controls to reduce their exposure to abusive interactions34. In January 2022, Instagram explained that it removes content that violates its and would show posts suspected of containing bullying, hate speech or incitement to violence lower in followers’ Feed and Stories5.

What Instagram prohibits and how violations are enforced

Instagram’s hate-speech rules also prohibit attacks based on ethnicity, national origin, sex, gender, gender identity, sexual orientation, disability or disease6. Mosseri explained that these protections cover attacks on whole groups as well as individuals, and that using a slur can cross the boundary7. He also distinguished removing a post from suspending an account: one removal does not automatically cause suspension, but accumulated violations, particularly serious ones, can lead to it8.

How hate speech is detected and measured

Meta research and patents. These may not describe how Instagram works today.

For suspected hate speech found outside private messages, Meta’s Newsroom says its technology can send content to human reviewers or delete it automatically when sufficiently confident that it violates the rules2. For January through March 2021, Meta’s Newsroom reported taking action on 6.3 million pieces of hate speech on Instagram, with 93% found before anyone reported them; these figures distinguish the amount acted on from the share detected before reports1. For detecting hate speech in recordings, a Meta patent describes a possible system that analyzes audio using speech recognition, but the patent does not establish that Instagram uses that system9.

How it has changed

  1. 2018

    April

    Instagram published FAQs setting out its prohibition on hate speech6.

  2. 2019

    July

    Instagram announced that accounts could be disabled for a certain number of violations within a time window, alongside its existing percentage-based threshold, and introduced notifications about accounts at risk10.

  3. 2020

    September

    Meta’s Newsroom announced policy updates explicitly covering certain implicit forms of hate speech, including blackface and stereotypes about Jewish people11.

  4. 2021

    May

    Meta’s Newsroom reported introducing stricter penalties as part of its work against hate speech and abuse1.

  5. July

    Mosseri said Instagram was testing 3.

  6. August

    Meta’s Newsroom announced , stronger offensive-comment warnings and the global rollout of on Instagram4.

  7. 2022

    January

    Instagram announced that posts suspected of containing hate speech would appear lower in their creators’ followers’ Feed and Stories5.

Controls for limiting exposure to abuse

Instagram’s controls against hateful interactions include comment filtering that lets people choose particular keywords to block, as Mosseri explained3. filters abusive direct-message requests, extending protection beyond public comments34. Limits lets people restrict comments and direct-message requests during spikes of attention4. Stronger warnings also intervene when someone tries to post a potentially offensive comment, before it is published4.

How hate speech affects recommendations and features

Instagram distinguishes removing hate speech under its from deciding what it can suggest to people in Explore and Reels under its 512. Its announced Link sticker policy also excludes accounts that repeatedly share hate speech or other rule-breaking content, as well as new accounts13. Instagram directs people who believe a restriction is mistaken to its information about account status and appeals13.

See also

References

  1. 1Keeping Athletes Safe This SummerMeta Newsroom · May 26, 2021
  2. 2Protecting Footballers and Fans on Our Apps During the FIFA World CupMeta Newsroom · Nov 17, 2022
  3. 3Racism is not acceptable and has no place on Instagram.Adam Mosseri · Jul 22, 2021
  4. 4Protecting Our Community from Abuse on InstagramMeta Newsroom · Aug 11, 2021
  5. 5How We Address Potentially Harmful Content on Feed and StoriesInstagram explainer · Jan 20, 2022
  6. 6Instagram Community Guidelines FAQsInstagram announcement · Apr 19, 2018
  7. 7Benny is a hysterical and prolific creator.Adam Mosseri · Feb 19, 2021
  8. 8@mosheh is a great news creator on Instagram, and my wife’s primary source of news these days.Adam Mosseri · Jul 27, 2021
  9. 9Classifiers for media contentMeta patent · Mar 22, 2022
  10. 10Account Disable Policy Changes on InstagramInstagram announcement · Jul 18, 2019
  11. 11An Update on Our Equity WorkMeta Newsroom · Sep 9, 2020
  12. 12How the “Algorithm” WorksAdam Mosseri · Jun 23, 2021
  13. 13Expanding Sharing Links in Stories to EveryoneInstagram announcement · Oct 27, 2021

Follow changes to Hate speech

Unofficial. Not affiliated with Instagram or Meta. AI answers from the sources shown.