By Brett Wilkins

I’m a journalist, so you can be assured that if I post a news link on Facebook, it’s legit. No fake news here. And so you can imagine my reaction when I checked my Facebook notifications Tuesday afternoon and saw that almost every one of my COVID-19-related posts over the past few days had been erroneously flagged and blocked.

“Your post goes against our Community Standards on spam,” a message in my support inbox explained. “No one else can see your post. We have these standards to prevent things like false advertising, fraud and security breaches.”

Naturally, I took to Facebook to sound the alarm. I found that I was far from the only one experiencing what was being described by my more conspiracy-minded Facebook friends as intentional censorship, and by my more tech-minded ones as bots gone wild in the absence of human content moderators, who are home locked down in the face of the rapidly-spreading coronavirus pandemic.

The latter seem to have been (mostly) right. Guy Rosen, Facebook’s vice president of integrity (hold that thought; this isn’t the right article for it), quickly tweeted that the problem “is a bug in an anti-spam system,” but was “unrelated to any change in our content moderator workforce.”

However, many people questioned how it is possible that such changes — Facebook has indefinitely sent home all content moderators, with pay — could not affect the quality of machine-driven moderation. The company had already admitted as much. “We may see some longer response times and make more mistakes as a result,” it said in a warning on Monday.

Facebook wasn’t alone. Alphabet Inc.’s YouTube, Twitter and Google all warned earlier this week that content could be wrongfully removed for policy violations. “Our automated systems may not always accurately classify content for removal, and human review of these decisions may be slower,” Google warned in a blog post.

As Facebook scrambled to fix the bug, many users noted that it wasn’t just COVID-19 posts that were getting flagged and blocked, while others complained of inaction when reporting actual fraudulent ads and spam. Some stories were truly heartbreaking. “Facebook flagged my GoFundMe post to raise funds for burial expenses for my great uncle who passed away unexpectedly on Sunday,” a user named Jennifer Hernandez wrote.

At 6:31 pm on Tuesday Rosen tweeted that Facebook had “restored all the posts that were incorrectly removed, which included posts on all topics, not just those related to COVID-19.” He added that “this was an issue with an automated system that removes links to abusive websites, but incorrectly removed a lot of other posts too.”

As soon as Rosen pressed “tweet” people began tweeting back that he was lying, that posts hadn’t been restored and that posts were still being flagged and blocked. It’s now Wednesday morning and all of my affected posts seem to be just fine. In a Wednesday call with reporters, Facebook CEO Mark Zuckerberg confirmed the widespread false positives but denied that they had anything to do with the fact that content moderators were working from home.

“This was just a technical error,” Zuckerberg said. “This was not because of coronavirus, and it was not because of our change in approach to… content moderation strategy.”


0 Comments

Leave a Reply

Your email address will not be published. Required fields are marked *

Please subscribe to receive the latest newsletter!
We respect your privacy.