Facebook is currently testing a new feature for Messenger that would identify suspicious accounts sending unwarranted direct messages, letting users know the account’s country of origin by phone number and whether it was recently created. The new feature was first disclosed by Motherboard, which received a screenshot of the test from multimedia artist Erin Gallagher. Facebook confirmed to Motherboard that it is indeed testing the feature, but spokesperson Dalya Browne said it was a “small test” at the moment.
Though the feature isn’t specifically aimed at Russian bot and scam accounts, the provided screenshot of the test indicates that it can notify a user when an unsolicited message was sent from a Messenger account that’s not paired with an official Facebook one and when that account is tied to a Russian phone number. The screenshot also shows how Facebook will list the account as “recently created.” Implicitly, Russia is Facebook’s primary adversary in its fight against misinformation and fake news.
The social network is still dealing with the aftermath of Russia’s massive state-sponsored misinformation campaign conducted in the run up to, during, and after the 2016 presidential election. The company is still actively suspending accounts tied to the Internet Research Agency, a Kremlin-tied propaganda outfit dedicated to seeding disinformation and distrust in the America public by way of fake accounts, pages, news, and comments on US social networking sites. Facebook is also now working with researchers to measure the impact of fake news on democratic elections and the overall effect of misinformation in the News Feed.
In addition to those efforts, Facebook also launched a new literacy campaign in May alongside a massive print advertising campaign to help educate both the general public and its active user base about misinformation and how to spot and stop the spread of fake news. The company is committing resources in its artificial intelligence research division to using automated software to fight misinformation, though that broader AI-driven moderation effort has significant roadblocks to overcome to successfully function as an alternative to human-guided moderation and direct oversight.