Media & Entertainment

Facebook’s ad system shown failing to enforce its own anti-discriminatory policy

Comment

The Facebook thumbs down image.
Image Credits: Bryce Durbin / TechCrunch

Can Facebook be trusted to abide by even its own stated standards? In the case of Internet political advertising the social giant wants to be allowed to continue to self regulate — despite the scandal of Russian bought socially divisive ads which (we now know) were tainting democratic discussion during the 2016 US presidential election (and beyond).

Don’t regulate us, we can regulate ourselves — honest!‘ is shaping up to be CEO Mark Zuckerberg’s massively moonshot new year project for 2018.

But results from a new ProPublica investigation suggest the tech giant is failing at even simple self-policing — undermining any claims it can responsibly manage the bad and even out-and-out illegal outcomes that are being enabled via its platform, and bolstering the case for more formal regulation.

Case in point: A year ago Facebook said it would disable ethnic affinity ad targeting for housing, employment and credit-related ads, following a ProPublica investigation that had suggested the platform’s ad-targeting capabilities could be used for discriminatory advertising — particularly in housing and employment, where such practices are illegal.

This month ProPublica checked in again, to see how Facebook is doing — by purchasing dozens of rental housing ads and asking that Facebook’s ad platform exclude groups that are protected from discrimination under the US Federal Fair Housing Act — such as African Americans and Jews.

Its test ads promoted a fictional apartment for rent, targeted at people aged 18 to 65 who were living in New York, house hunting and likely to move — with ProPublica narrowing the audience by excluding certain “Behaviors”, listed in a section Facebook now calls “Multicultural Affinity”, including “Hispanic”, “African American” and “Asian American”.

However instead of the platform blocking the potentially discriminatory ad buys, ProPublica reports that all its ads were approved by Facebook “within minutes” — including an ad that sought to exclude potential renters “interested in Islam, Sunni Islam and Shia Islam”. It says that ad took the longest to approve of all its buys (22 minutes) — but that all the rest were approved within three minutes.

It also successfully bought ads that it judged Facebook’s system should at least flag for self-certification because they were seeking to exclude other members of protected categories. But the platform just accepted housing ads blocked from being shown to categories including ‘soccer moms’, people interested in American sign language, gay men and people interested in wheelchair ramps.

Yet, back in February, Facebook announced new “stronger” anti-discriminatory ad polices, saying it was deploying machine learning tech tools to help it identify ads in the categories of concern.

“We’ve updated our policies to make our existing prohibition against discrimination even stronger. We make it clear that advertisers may not discriminate against people based on personal attributes such as race, ethnicity, color, national origin, religion, age, sex, sexual orientation, gender identity, family status, disability, medical or genetic condition,” it wrote then.

Of the new tech tools, Facebook said: “This will allow us to more quickly provide notices and educational information to advertisers — and more quickly respond to violations of our policy.”

Explaining how the new system would work, Facebook said advertisers who attempt to show “an ad that we identify as offering a housing, employment or credit opportunity” and which “either includes or excludes our multicultural advertising segments — which consist of people interested in seeing content related to the African American, Asian American and US Hispanic communities” will find the platform disapproves the ad.

The new system would also require all advertisers that attempt to buy targeted advertising in the categories of concern to self-certify they are complying with Facebook’s anti-discrimination policies and with “applicable anti-discrimination laws”.

ProPublica says it never even encountered these self-certification screens, as well as never having any of its ad buys blocked.

“Under its own policies, Facebook should have flagged these ads, and prevented the posting of some of them. Its failure to do so revives questions about whether the company is in compliance with federal fair housing rules, as well as about its ability and commitment to police discriminatory advertising on the world’s largest social network,” it writes.

Responding to ProPublica’s findings, Facebook sent a statement attributed to Ami Vora, VP of product management, in which she concedes its system failed in this instance. “This was a failure in our enforcement and we’re disappointed that we fell short of our commitments. The rental housing ads purchased by ProPublica should have but did not trigger the extra review and certifications we put in place due to a technical failure,” said Vora.

She went on to claim Facebook’s anti-discrimination system had “successfully flagged millions of ads” in the credit, employment and housing categories — but also said Facebook will now begin requiring self-certification for ads in all categories that choose to exclude an audience segment.

“Our systems continue to improve but we can do better,” she added.

The latter phrase is now a very familiar refrain from Facebook where content review and moderation is concerned. Aside from socially divisive political disinformation, it has faced growing criticism this year for enabling the spread of content such as extremist propaganda and child exploitation, as well as for multiple incidents of its tools being used to broadcast suicides and murders.

The wider question for governments and regulators is at what point will Facebook’s attempts to ‘do better’ be deemed just not good enough?

Commenting on ProPublica’s findings in a statement, Rachel Goodman, an attorney with the ACLU‘s Racial Justice Program, said: “We’re very, very disappointed to see these significant failures in Facebook’s system for identifying and preventing discrimination in advertisements for rental housing. We and other advocates spent many hours helping Facebook move toward solving the egregious discrimination problem built into its ad targeting business — that advertisers could exclude people from seeing ads based on race, gender, and religion, including in ads for housing, credit, and employment. Facebook’s representations to us indicated that this problem had been substantially solved, but it now seems clear that was not the case.

“While we appreciate that Facebook continues to express a desire to get it right on this important civil rights issue, this story highlights the need for greater transparency and accountability. Had outside researchers been able to see and the system Facebook created to catch these ads, those researchers could have spotted this problem and ended the mechanism for discrimination sooner.”

This story was updated with additional comment from the ACLU

More TechCrunch

For over six decades, the nonprofit has been active in the financial services sector.

Accion’s new $152.5M fund will back financial institutions serving small businesses globally

Meta’s newest social network, Threads is starting its own fact-checking program after piggybacking on Instagram and Facebook’s network for a few months. Instagram head Adam Mosseri noted that the company…

Threads finally starts its own fact-checking program

Looking Glass makes trippy-looking mixed-reality screens that make things look 3D without the need of special glasses. Today, it launches a pair of new displays, including a 16-inch mode that…

Looking Glass launches new 3D displays

Replacing Sutskever is Jakub Pachocki, OpenAI’s director of research.

Ilya Sutskever, OpenAI co-founder and longtime chief scientist, departs

Intuitive Machines made history when it became the first private company to land a spacecraft on the moon, so it makes sense to adapt that tech for Mars.

Intuitive Machines wants to help NASA return samples from Mars

As Google revamps itself for the AI era, offering AI overviews within its search results, the company is introducing a new way to filter for just text-based links. With the…

Google adds ‘Web’ search filter for showing old-school text links as AI rolls out

Blue Origin’s New Shepard rocket will take a crew to suborbital space for the first time in nearly two years later this month, the company announced on Tuesday.  The NS-25…

Blue Origin to resume crewed New Shepard launches on May 19

This will enable developers to use the on-device model to power their own AI features.

Google is building its Gemini Nano AI model into Chrome on the desktop

It ran 110 minutes, but Google managed to reference AI a whopping 121 times during Google I/O 2024 (by its own count). CEO Sundar Pichai referenced the figure to wrap…

Google mentioned ‘AI’ 120+ times during its I/O keynote

Firebase Genkit is an open source framework that enables developers to quickly build AI into new and existing applications.

Google launches Firebase Genkit, a new open source framework for building AI-powered apps

In the coming months, Google says it will open up the Gemini Nano model to more developers.

Patreon and Grammarly are already experimenting with Gemini Nano, says Google

As part of the update, Reddit also launched a dedicated AMA tab within the web post composer.

Reddit introduces new tools for ‘Ask Me Anything,’ its Q&A feature

Here are quick hits of the biggest news from the keynote as they are announced.

Google I/O 2024: Here’s everything Google just announced

LearnLM is already powering features across Google products, including in YouTube, Google’s Gemini apps, Google Search and Google Classroom.

LearnLM is Google’s new family of AI models for education

The official launch comes almost a year after YouTube began experimenting with AI-generated quizzes on its mobile app. 

Google is bringing AI-generated quizzes to academic videos on YouTube

Around 550 employees across autonomous vehicle company Motional have been laid off, according to information taken from WARN notice filings and sources at the company.  Earlier this week, TechCrunch reported…

Motional cut about 550 employees, around 40%, in recent restructuring, sources say

The keynote kicks off at 10 a.m. PT on Tuesday and will offer glimpses into the latest versions of Android, Wear OS and Android TV.

Google I/O 2024: Watch all of the AI, Android reveals

Google Play has a new discovery feature for apps, new ways to acquire users, updates to Play Points, and other enhancements to developer-facing tools.

Google Play preps a new full-screen app discovery feature and adds more developer tools

Soon, Android users will be able to drag and drop AI-generated images directly into their Gmail, Google Messages and other apps.

Gemini on Android becomes more capable and works with Gmail, Messages, YouTube and more

Veo can capture different visual and cinematic styles, including shots of landscapes and timelapses, and make edits and adjustments to already-generated footage.

Google Veo, a serious swing at AI-generated video, debuts at Google I/O 2024

In addition to the body of the emails themselves, the feature will also be able to analyze attachments, like PDFs.

Gemini comes to Gmail to summarize, draft emails, and more

The summaries are created based on Gemini’s analysis of insights from Google Maps’ community of more than 300 million contributors.

Google is bringing Gemini capabilities to Google Maps Platform

Google says that over 100,000 developers already tried the service.

Project IDX, Google’s next-gen IDE, is now in open beta

The system effectively listens for “conversation patterns commonly associated with scams” in-real time. 

Google will use Gemini to detect scams during calls

The standard Gemma models were only available in 2 billion and 7 billion parameter versions, making this quite a step up.

Google announces Gemma 2, a 27B-parameter version of its open model, launching in June

This is a great example of a company using generative AI to open its software to more users.

Google TalkBack will use Gemini to describe images for blind people

Google’s Circle to Search feature will now be able to solve more complex problems across psychics and math word problems. 

Circle to Search is now a better homework helper

People can now search using a video they upload combined with a text query to get an AI overview of the answers they need.

Google experiments with using video to search, thanks to Gemini AI

A search results page based on generative AI as its ranking mechanism will have wide-reaching consequences for online publishers.

Google will soon start using GenAI to organize some search results pages

Google has built a custom Gemini model for search to combine real-time information, Google’s ranking, long context and multimodal features.

Google is adding more AI to its search results