By Sarah E. Needleman and Jeff Horwitz 

Civil-rights advocates are increasing pressure on Facebook Inc. advertisers to halt spending on the company's platforms, saying it has done too little to police hateful and other problematic content.

Groups including the Anti-Defamation League and the NAACP have enlisted hundreds of companies in the boycott campaign, while lawyers Facebook hired to audit its handling of civil-rights issues issued a report this week saying the company made progress on several fronts but has more work to do.

Facebook has acknowledged its shortcomings and pointed to new policies, additional spending and other efforts aimed at trying to address concerns.

Here is a look at what Facebook is doing -- and not doing -- in response.

What specific demands are civil-rights groups making of Facebook?

Stop Hate for Profit, the name of the coalition behind the boycott effort, has on its website 10 recommendations for Facebook, ranging from the specific to the sweeping.

The group says Facebook should hire an executive with civil-rights expertise for a post in its C-suite, refund advertisers whose ads were shown next to content later removed for violating Facebook's terms of service and submit to regular independent audits of identity-based hate and misinformation on its platform.

On Wednesday, Facebook said it plans to hire a vice president for civil rights, a role that a spokeswoman described as "very senior" and is intended to lead a team built out over time. She also said the company issues refunds to advertisers in some instances.

The boycott groups credit Facebook with some progress on those points but say it is insufficient. The coalition wants Facebook to adopt "common-sense changes" to its policies to reduce hate on the platform and remove content that could inspire people to commit violence; stop amplifying all groups associated with hate, misinformation or conspiracies; and remove groups focused on "white supremacy, militia, anti-Semitism, violent conspiracies, Holocaust denialism, vaccine misinformation and climate denialism."

How does Facebook define hate speech?

Facebook's community standards define hate speech as attacks on people based on about a dozen "protected characteristics" including race, religious affiliation, national origin and gender identity. Attacks include violent or dehumanizing speech, statements of inferiority or calls for exclusion or segregation, it says.

Facebook has broadened its definition of hate speech and who is protected over time. Last month, it said it would stop allowing ads alleging that a particular race, religion or identity group posed a threat to others. Such content is still allowed in unpaid posts.

Civil-rights advocates say the rules remain too narrow, allowing white supremacists to dodge crackdowns by avoiding certain keywords that would surface in searches for hate speech. The Anti-Defamation League last month listed several examples of hateful or extremist posts on Facebook that still appear near major companies' ads on the platform.

One example is of an ad from home-sharing company Airbnb Inc. that the League said appeared next to a post from an antigovernment militia movement called the Three Percenters, about the recent decision by PepsiCo Inc.'s Quaker Oats brand to retire its Aunt Jemima branding. In another, the League said an ad from human-resources company Randstad Holding NV appeared next to a video from a large Facebook group called Q-Anon Patriots that accused the media of promoting cannibalism by attempting to draw a connection between a sculpture owned by former lobbyist Tony Podesta and deceased serial killer Jeffrey Dahmer.

The Facebook spokeswoman said it takes aggressive action against groups and people who promote hate. The company last month said it banned hundreds of accounts deemed to have links to white supremacist organizations or groups that promote violence.

How does Facebook handle speech by political figures compared with everyday users?

Some of the biggest recent battles over content have involved President Trump's social-media posts.

When he wrote "when the looting starts, the shooting starts" in response to protests and unrest after the police killing of George Floyd, Twitter Inc. labeled the post as violating its policies on glorifying violence; Facebook left the same comments untouched.

Mr. Trump responded that he wasn't urging police to shoot protesters, adding that "nobody should have any problem with this other than the haters, and those looking to cause trouble on social media."

Facebook has said it doesn't fact-check political speech from politicians because it doesn't believe private companies should be arbiters of what is true and what isn't.

The boycott coalition says Facebook should eliminate its exemption for politicians.

Last month CEO Mark Zuckerberg said Facebook would start labeling posts by any user that violate content rules but are deemed newsworthy. For example, the company determined that posts showing a 1972 war photo of a naked Vietnamese girl fleeing napalm bombs was newsworthy and could remain on the platform, after Facebook initially took down that content under restrictions on child nudity.

Mr. Zuckerberg also said Facebook would add safeguards to prevent content that promotes voter suppression and ads that depict immigrants or asylum seekers as inferior. Facebook would have to decide whether nonadvertising content that is reported as abusive toward immigrants should be removed.

What tools does Facebook use to moderate content on its platform?

Facebook says it has more than 35,000 people working on safety and security, including more than 350 employees with expertise in law enforcement, national security, counterterrorism intelligence and academic studies in radicalization. Facebook also uses artificial intelligence to detect prohibited content, and says it removes most of that content before there is a user report. The company says it blocks millions of fake accounts from being created every day.

The amount of effort Facebook puts into detecting violations is a point of contention. The civil-rights groups say the company has enormous financial resources but still routinely fails to catch toxic content posted within group pages and elsewhere. They say all individuals facing severe hate and harassment should be able to connect with a live Facebook employee.

The Facebook spokeswoman said users can moderate comments on their posts, block other users and control their posts' visibility by creating a restricted list. She also said the company made changes allowing groups to be removed if an administrator encourages or creates posts that violate platform rules.

To help address where lines should be drawn on hate speech, Facebook has created an independent content governance board that includes human-rights lawyers and free-speech advocates. The group likely will be fully operational late this year.

What is Facebook's 'transparency' report?

The report created by Facebook shows how the company enforces its community standards and content restrictions, responds to data requests and protects intellectual property. The company's most recent transparency report, released in May, says Facebook removed 9.6 million pieces of content with hate speech in the first quarter, up from 5.7 million in the fourth quarter of 2019.

The civil-rights groups say they don't trust Facebook to check its own work. They say its report doesn't address hate speech that isn't reported or whether it dismisses concerns about content that is reported as abusive but isn't removed.

"A 'transparency report' is only as good as its author is independent," the Stop Hate for Profit website says.

The Facebook spokeswoman said the company has committed to providing more insight into how it enforces its hate-speech policies in the coming year.

Does Facebook remove whole groups from its platform?

Facebook has struggled to articulate exactly when a private group becomes a threat to other users, but it is a sliding scale. Groups that Facebook considers spreaders of misinformation or vitriolic content can be removed from its algorithmic recommendation systems. Permanent removal is a possibility for entities deemed a threat to others, but Mr. Zuckerberg has voiced reluctance to entirely prevent users from forming communities around subjects they care about.

Facebook also removes groups and pages it identifies as part of coordinated inauthentic behavior.

Who is ultimately in charge of what types of content are allowed on the company's social-media platforms?

Mr. Zuckerberg has final say on content policy matters. The CEO guided Facebook's handling of the president's social-media posts -- a focus of its dispute with civil-rights groups.

Mr. Zuckerberg has argued for free expression online, saying in a speech at Georgetown University in October that Facebook shouldn't be the ultimate arbiter of truth. "In a democracy, I believe people should decide what is credible, not tech companies," he said. Mr. Zuckerberg also said creating rules that prohibit free speech can have unintended consequences.

The Facebook spokeswoman said that the company makes tens of thousands of decisions about content daily and that the process depends on its roughly 15,000 content reviewers world-wide. In some cases, content decisions are escalated to people on different teams, and they are handled routinely by on-call content policy specialists around the world, with senior leadership weighing in on a small number of decisions, she said.

Write to Sarah E. Needleman at sarah.needleman@wsj.com and Jeff Horwitz at Jeff.Horwitz@wsj.com

 

(END) Dow Jones Newswires

July 09, 2020 14:34 ET (18:34 GMT)

Copyright (c) 2020 Dow Jones & Company, Inc.
Facebook (NASDAQ:FB)
Historical Stock Chart
From Jul 2020 to Aug 2020 Click Here for more Facebook Charts.
Facebook (NASDAQ:FB)
Historical Stock Chart
From Aug 2019 to Aug 2020 Click Here for more Facebook Charts.