Skip to main content

Facebook to start testing AI review system for discriminatory advertising

faceook ad policies discrimination facebook ads
Facebook is continuing its efforts to ensure nobody uses its tools to create biased ads that exclude people based on a number of factors, such as race, ethnicity, age, or sexual orientation.

On Wednesday, the company announced it is updating its ad policies to make its existing prohibiting of discrimination even stronger. Additionally, it is testing an automated system that uses machine learning technology to identify potentially controversial ads. Facebook claims the new model will help it to streamline the review process.

The social network has also added a new section linked from its ad policies page that provides educational resources from government agencies and civil rights groups, including the ACLU and Department of Justice.

Facebook’s comprehensive ad targeting tools came under fire last year when it was alleged that they included a feature that was unlawful. The function in question allowed marketers to exclude users from seeing their ads based on ethnicity. Investigative news nonprofit ProPublica claimed the option was in violation of federal housing and employment laws, which prohibit ads that discriminate against people based on such factors as race and gender. The article was followed by a lawsuit (filed by several Facebook users) that made similar allegations.

Facebook responded by swiftly disabling the “ethnic affinity” feature for ads pertaining to housing, employment, or credit. Now, when someone tries to create a Facebook ad related to those fields and attempts to excludes the site’s multicultural advertising segments (African-American, Asian-American and U.S. Hispanic communities) the ad is disapproved. In its warning message to advertisers trying to create these types of items, Facebook allows the creator to submit the ad for a manual review.

Similarly, if an advertiser tries to show a housing, employment, or credit ad that uses any other audience segment, they will be required to certify that they are complying with the new policy and with applicable anti-discrimination laws.

“Discriminatory advertising has no place on Facebook,” states the company in a blog post. “Over the past several months, we’ve met with policymakers and civil rights leaders to gather feedback about ways to improve our enforcement while preserving the beneficial uses of our advertising tools.”

Editors' Recommendations

Saqib Shah
Former Digital Trends Contributor
Saqib Shah is a Twitter addict and film fan with an obsessive interest in pop culture trends. In his spare time he can be…
Amazon sues 10,000 Facebook groups over fake reviews
Amazon logo on the headquarters building.

When it comes to reviews on Amazon and similar shopping sites, most people have by now developed their own approach to dealing with them.

Some use a blend of instinct and experience to decide if what they’re reading is genuine, while others scan a broad selection to try to get an overall feel for a product’s reputation. Of course, some folks simply ignore them altogether.

Read more
Your Facebook account could get multiple profiles for different interests
The Facebook home page on a screen.

Facebook is testing a way to give its users more profiles per account, ostensibly to give users more opportunities for sharing posts and keeping up with the platform's content.

On Thursday, Bloomberg reported that Meta (Facebook's parent company) would begin experimenting with letting some Facebook users generate up to four other profiles in addition to their main account's profile.

Read more
Meta’s new AI research may boost translations on Facebook, Instagram
Image with languages displaying in front of a man on his laptop for Meta's 200 languages within a single AI model video.

Facebook's parent company, Meta, announced a new AI model today that can translate hundreds of languages, and its research is expected to help improve language translations on its social media apps, specifically Facebook and Instagram.

On Wednesday, Meta unveiled its new AI model, NLLB-200. NLLB stands for No Language Left Behind, which is a Meta project that endeavors to "develop high-quality machine translation capabilities for most of the world’s languages." The AI model that came from that project, NLLB-200, can translate 200 languages.

Read more