A man holds a smart phone with the icons for the social networking apps Facebook, Instagram and Twitter seen on the screen in Moscow on March 23, 2018.
A public apology by Facebook chief Mark Zuckerberg, on March 22, 2018 failed to quell outrage over the hijacking of personal data from millions of people, as critics demanded the social media giant go much further to protect privacy. / AFP PHOTO / Kirill KUDRYAVTSEV        (Photo credit should read KIRILL KUDRYAVTSEV/AFP/Getty Images)
Lies, smears and scams plaguing social media
07:09 - Source: CNN
New York CNN Business  — 

Facebook and its CEO, Mark Zuckerberg, pledged to fix how it handles political and issue ads in the wake of Russian meddling in 2016. But just days before the midterm elections, a key part of Facebook’s effort is broken, and it’s unclear if the company is doing anything to fix it.

The company has touted new rules for political ad-buyers as a major component of its work to combat disinformation on its platform. In 2016 Russian trolls with links to the Kremlin bought ads targeting Americans in the run-up to the presidential election. They were able to do so without giving any information to Americans seeing those ads about who was paying for them.

Political ads on the platform are now supposed to say who paid for them, but Facebook allows buyers to fill in that information themselves. And if anyone or any system at the company is supposed to be ensuring that the information these ad-buyers submit is the truth, they appear to be asleep at the wheel.

Earlier this week, Vice News, posing as a political ad-buyer, got approval from Facebook to run ads in the name of every single one of the US’ 100 senators. Vice News did not end up buying the ads. This came after Vice News had previously received approval from Facebook to run ads “Paid for” by Islamic State and Vice President Mike Pence.

On Thursday, ProPublica reported it had identified issues with how private interest groups use the Facebook disclaimer.

Separately, CNN Business found one advertiser breaking Facebook’s rules twice in one week. Facebook missed both violations and only acted in both instances after CNN inquired about it.

The advertiser CNN Business found, an anti-Ted Cruz page called “Crush Cruz,” contained no information about who was behind it. Its ad disclaimers simply read, “Paid for by Crush Cruz.” The page had spent almost $6,000 on ads, meaning it could have reached hundreds of thousands, possibly millions, of Texas voters.

Last Thursday after CNN Business asked about the page, Facebook said that Crush Cruz’s disclaimer did not accurately represent a person or entity and stopped it from running ads.

The next day, the page was running ads again. This time, its disclaimer read, “Paid for by Crush Cruz. Not authorized or paid for by any PAC, candidate, or candidate’s committee.”

Facebook allowed the page to run 14 ads with the new disclaimer between Friday and Tuesday, when CNN Business contacted the company about the page again.

Once again, Facebook stopped the page’s ads — saying the new disclaimer also broke its policies.

But why did Facebook not more carefully review a disclaimer submitted by a page that had already broken its rules? How did it not notice that it was going to approve ads falsely claiming to be run by Pence, 100 senators and the Islamic State.

It’s not saying.

CNN Business asked the company if a human reviewed Crush Cruz’s new disclaimer before the page was allowed to begin running ads again, and if it were Facebook’s policy to have a human review disclaimers for a page it had previously determined was in violation of its rules.

The company would only say its reviews are a combination of automation and human, and would provide no further details on any process it has for humans to vet the disclaimers.

Rob Leathern, director, product management at Facebook, told CNN Business in a statement, “We believe that ad transparency is key to helping prevent election interference. Inaccurate ‘paid for by’ labels violate our policies. On top of our existing proactive checks, when alerted to deceptive labels, we investigate and remove them.”

Leathern said that the company’s new publicly-available ad archive allows news organizations to hold bad actors and Facebook itself accountable.

He said, “It’s a way we’re also held accountable– even if it means our mistakes are on display. We’re exploring additional checks to help prevent abuse and will respond to requests from law enforcement and election officials now and in the future if new requirements arise.”

Editor’s note: Facebook has given CNN and other organizations access to its ad archive API to analyze political ad spending. The data featured in this story was gathered through Facebook’s public political ad website, not its ad archive API.