Facebook’s automated moderation tools went wild and targeted tons of posts about the coronavirus pandemic and other topics on Tuesday evening, blocking users from sharing articles from legitimate news sources.
After former Facebook chief security officer Alex Stamos tweeted that the errors could be “the start of the [machine learning] going nuts with less human oversight,” current vice president of integrity Guy Rosen responded that “a bug in an anti-spam system” was to blame. He also shot down speculation that the problem was related to an outbreak-related shortage of onsite human moderators. (Facebook said it would pay full time working from home $US1,000 ($1,668) to deal with emergency conditions, but the long-suffering contract workers that staff its moderation systems are only eligible for continued pay.)
“We’re on this—this is a bug in an anti-spam system, unrelated to any changes in our content moderator workforce,” Rosen wrote. “We’re in the process of fixing and bringing all these posts back.” Later on Tuesday evening, he tweeted that the issue was not specific to coronavirus-related content, adding that it should be fixed and all posts erroneously flagged as spam should be restored. However, other Twitter users responded that the issues weren’t yet resolved for them.
Screenshots posted to Twitter and elsewhere showed the system was informing users, “Your post goes against our Community Standards on spam.” Facebook had previously conceded that the spread of coronavirus throughout the U.S. might tax the abilities of its automated tools to the limit.
.@Facebook — Oh wow and then three in a row !
How is @TomHanks being okay be SPAM? And again this is from @Upworthy not exactly a new publication. All I did was share it. pic.twitter.com/SYhYknzJ9Z
— Kristine Schachinger (@schachin) March 17, 2020
Yes! Including one about horoscopes pic.twitter.com/IQCkQVNPhT
— Gina G (@AnotherGinaG) March 18, 2020
As TechCrunch noted, Facebook’s sheer scale means that it plays a central role in transmitting information about the outbreak. Breakdowns in automated systems could thus have a greater impact than during business as usual, with major disruptions to daily life and concern that the situation could grow worse playing out across the globe.
Facebook has claimed to be fighting misinformation on its platform about the virus and the disease it causes, covid-19, as well as to be restricting ads for products that fraudulently claim to cure it. It’s also placed a ban on advertising face masks after shortages resulted in widespread price gouging. This week, it signed a joint pledge to prevent the spread of false information with Google, LinkedIn, Microsoft, Reddit, Twitter, YouTube, but Facebook’s shoddy track record on effectively moderating its platform isn’t exactly a great omen for how it will handle the flood of coronavirus B.S. already spewing out.