{"id":94724,"date":"2019-08-13T07:02:26","date_gmt":"2019-08-13T14:02:26","guid":{"rendered":"https:\/\/lifeboat.com\/blog\/2019\/08\/googles-algorithm-for-detecting-hate-speech-is-racially-biased"},"modified":"2019-08-13T07:02:26","modified_gmt":"2019-08-13T14:02:26","slug":"googles-algorithm-for-detecting-hate-speech-is-racially-biased","status":"publish","type":"post","link":"https:\/\/lifeboat.com\/blog\/2019\/08\/googles-algorithm-for-detecting-hate-speech-is-racially-biased","title":{"rendered":"Google\u2019s algorithm for detecting hate speech is racially biased"},"content":{"rendered":"

<\/a><\/p>\n

Algorithms meant to spot hate speech online are far more likely to label tweets \u201coffensive\u201d if they were posted by people who identify as African-American.<\/p>\n

\n

AI systems meant to spot abusive online content are far more likely to label tweets \u201coffensive\u201d if they were posted by people who identify as African-American.<\/p>\n

The news:<\/strong> Researchers built two AI systems<\/a> and tested them on a pair of data sets of more than 100,000 tweets that had been annotated by humans with labels like \u201coffensive,\u201d \u201cnone,\u201d or \u201chate speech.\u201d One of the algorithms incorrectly flagged 46% of inoffensive tweets by African-American authors as offensive. Tests on bigger data sets, including one composed of 5.4 million tweets, found that posts by African-American authors were 1.5 times more likely to be labeled as offensive. When the researchers then tested Google\u2019s Perspective<\/a>, an AI tool that the company lets anyone use to moderate online discussions, they found similar racial biases.<\/p>\n

A hard balance to strike:<\/strong> Mass shootings perpetrated by white supremacists in the US and New Zealand have led to growing calls from politicians for social-media platforms to do more to weed out hate speech. These studies underline just how complicated a task that is. Whether language is offensive can depend on who\u2019s saying it, and who\u2019s hearing it. For example, a black person using the \u201cN word\u201d is very different from a white person using it. But AI systems do not, and currently cannot, understand that nuance.<\/p>\n","protected":false},"excerpt":{"rendered":"

Algorithms meant to spot hate speech online are far more likely to label tweets \u201coffensive\u201d if they were posted by people who identify as African-American. AI systems meant to spot abusive online content are far more likely to label tweets \u201coffensive\u201d if they were posted by people who identify as African-American. The news: Researchers built [\u2026]<\/p>\n","protected":false},"author":531,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[41,6],"tags":[],"class_list":["post-94724","post","type-post","status-publish","format-standard","hentry","category-information-science","category-robotics-ai"],"_links":{"self":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts\/94724","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/users\/531"}],"replies":[{"embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/comments?post=94724"}],"version-history":[{"count":0,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts\/94724\/revisions"}],"wp:attachment":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/media?parent=94724"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/categories?post=94724"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/tags?post=94724"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}