testsetset
In May, Facebook launched a political ad archive to make it easier for users to see the types of political ads being bought on the platform — even if those ads weren’t targeted at them.
Some advertisers have complained about Facebook’s implementation of the policy, as was to be expected. But within the past few weeks, a much bigger issue has bubbled to the surface — a loophole that undermines the ad archive’s stated goal of making it easier for users to understand who is paying for political ads.
In two different stories, Vice News reporters have detailed how they got approval to run ads on Facebook and were then able to write pretty much anything they wanted to in the “paid for by” label. (Political advertisers first have to get their identity and location verified by Facebook.)
Vice News got ads approved to run that were “paid for by”: Mike Pence, ISIS, and all 100 U.S. Senators.The only two names that got rejected? Mark Zuckerberg and Hillary Clinton.
June 5th: The AI Audit in NYC
Join us next week in NYC to engage with top executive leaders, delving into strategies for auditing AI models to ensure fairness, optimal performance, and ethical compliance across diverse organizations. Secure your attendance for this exclusive invite-only event.
Business Insider followed that up with a test of its own and found that Facebook gave it approval to run an ad “paid for by” the now-defunct Cambridge Analytica.
This egg on Facebook’s face comes less than a week before the U.S. midterms. This is the time Facebook was supposed to prove that it has learned from the lessons of 2016, when bad actors were able to exploit Facebook’s algorithms and lax sign-up requirements to spread misinformation through fake accounts.
“Inaccurate disclaimers have no place on Facebook, and these ads are no longer running,” Facebook’s Rob Leathern told Vice News in response to their first story. “Enforcement isn’t perfect — and we won’t stop all people trying to game the system — but we have made it much harder and we will continue to improve.”
Enforcement will never be perfect — but that doesn’t let Facebook off the hook for giving users the bare minimum information about who is paying for political ads in the first place.
I’ve written previously that Facebook’s ad archive is lacking in terms of the tools it gives users to explore the archive and link activity between entities. One way Facebook could limit the activity of bad actors trying to exploit this “paid for by” loophole is by giving users and researchers more information about political ads, making it easier to spot something fishy.
Damon McCoy, an assistant professor of computer science and engineering at New York University who has studied Facebook, Twitter, and Google’s new ad archives, told VentureBeat previously that he thinks Facebook should take a page from Google.
Google, in its political ad archive, lists the advertiser’s Employer Identification Number (EIN) or Federal Election Candidate (FEC) number. If it’s an FEC number, Google then links to that organization’s page on the FEC website. On the FEC website, users can see who gave money to a PAC or political organization. Facebook doesn‘t currently ask for an EIN or FEC number when verifying political advertisers.
But these numbers make it easier for users to get additional information about the organization in question and allow them to see if the number matches the name of the organization purportedly paying for the ad. Presumably, this would also make it easier for Facebook’s systems to catch discrepancies between the approved advertiser and the “paid for by” label.
This would also be more difficult for Facebook to implement, as the company includes issue ads in its archives — ads that take a stance on a political topic but don’t explicitly support a candidate. Ads bought by private companies often fall into this category. Google currently does not include issue ads in its archives.
It’s also worth noting that Facebook isn’t the only company letting political advertisers fill out the “paid for by” label by themselves. Twitter also lets advertisers do this; however, the company has thousands fewer political advertisers than Facebook, so it’s easier to catch deceptive activity. A Twitter spokesperson told VentureBeat that the company does monitor vetted political advertisers to ensure they don’t change the “paid for by” label to something deceptive.
Other ways Facebook could make it easier for users and researchers to dig into who is paying for political ads: Make the “paid for” label a linkable page so the user can see all of the ads purportedly paid for by the same organization. Or Facebook could allow the user to segment their searches within the political ad archive based on the “paid for by” label to see if a certain group is funding an unusual spectrum of ads.
Facebook also needs to add more reporting tools — currently, there’s no way for the average user to report a “paid for by” label that they think is suspicious.
Hopefully, this episode has finally convinced Facebook that just dumping a bunch of information — in this case, political ads — into one webpage and calling that an improvement is likely to backfire. If the company doesn’t give users, researchers, and its own employees more tools to spot deceptive activity on its platform it is going to continue to open itself up to a lot more embarrassing investigations at critical junctures, like this one.
