Facebook acknowledged Tuesday it has developed tools to identify users "indiscriminately" flagging fake news as it refines its effort to combat misinformation.

But the leading social network disputed as "just plain wrong" a Washington Post report that it has developed an overall "reputation score" for its users as part of the initiative.

Facebook said it has developed "a process to protect against people indiscriminately flagging news as fake and attempting to game the system" which relies in part on how often a user reports something as fake despite verification by fact-checkers.

"The reason we do this is to make sure that our fight against misinformation is as effective as possible," Facebook said in a statement.

Users who report what appears to be bogus news are given a standard probability score of from zero to one depending on how reliable they are when it comes to reporting posts that are untrue, according to the social network.

The rating is one of many "signals" used to prioritize flagged posts sent to be reviewed by fact-checking teams.

But Facebook said the Post report was misleading because it did not create a "unified score" to rank the overall trustworthiness of its users.

Over the past 18 months, Facebook and other online platforms have stepped up efforts to combat the spread of false news with the intent to manipulate the platforms.

Part of the challenge battling bogus content is that some people report posts as false simply because they disagree with stories, or in efforts to wrongly discredit them, according to the social network.

Repeatedly reporting accurate information to be false at Facebook would skew a users reliability rating toward zero in the ranking system.

Facebook last month shut down 32 fake pages and accounts involved in an apparent "coordinated" effort to stoke hot-button issues ahead of November midterm US elections.

The US intelligence community has concluded that Russia sought to sway the vote in Donald Trump's favor, and Facebook was a primary tool in that effort, using targeted ads to escalate political tensions and push divisive online content.

Facebook has since made a priority of preventing the social network to be used to spread misleading or outright deceitful messages aimed at influencing politics.

Facebook cuts ad-target options to thwart discrimination
San Francisco (AFP) Aug 21, 2018 –

Facebook said Tuesday is cutting more than 5,000 ad-targeting options to prevent advertisers from discriminating based on traits such as religion or race.

The shift eliminates the ability to direct Facebook ads at people based on ethnicity, beliefs, political affinity or other data that could be considered sensitive or personal.

"While these options have been used in legitimate ways to reach people interested in a certain product or service, we think minimizing the risk of abuse is more important," the leading online social network said in an online post.

The announcement came less than a week after the US Department of Housing and Urban Development accused Facebook of breaking the law by letting landlords and home sellers use its ad-targeting system to discriminate against potential buyers or tenants.

A formal complaint filed by HUD contended that Facebook advertisers were able to target offers of homes available for rent or sale based on factors such as race, religion, gender, nationality or disabilities.

"We're committed to protecting people from discriminatory advertising on our platforms," Facebook said.

Some media reports this year noted that advertisers could choose to target ads at African Americans, Hispanics or other demographic groups — or exclude them, to effectively market products or services to whites.

Most of the terms removed were in a category that let advertisers designate who they wanted excluded from seeing marketing messages.

For example, Facebook advertisers can no longer opt to prevent ads from being shown to user who have expressed interest in topics such as Passover, Islam, Buddhism or Native American culture.

Facebook added that all US advertisers will need to certify that they accept the social network's non-discrimination policy. Previously, the compliance certification was required only of those posting housing, employment or credit ads.

The certification is intended to educate advertisers regarding "the difference between acceptable ad targeting and ad discrimination," according to the social network.

Facebook planned to eventually expand the measure to more countries, but provided no timeline and insisted the move was not a reaction to the HUD complaint.

"No, we've been building these tools for a long time and collecting input from different outside groups," a Facebook spokesman said of the notion.

The HUD complaint filed on Friday came after an investigation confirmed that advertisers on Facebook could exclude categories such as people who expressed interest in assistance dogs, parenting, China or the Bible, according to the agency.

Facebook prohibits discrimination and has strengthened its systems during the past year to protect against advertisers misusing targeting capabilities, a spokesman told AFP.