Within the past 24 hours, Facebook was granted a patent for “Moderating content in an online forum,” which it first applied for in February of 2015. Unfortunately, the most accurate term to describe what it covers is shadowbanning.
Shadowbanning, which has become a favoured term among conservative and far-right interests who believe they’re experiencing ideological censorship at the hands of social networking platforms, describes a specific scenario wherein a user has their content demoted or outright hidden without their knowledge. Compare that vernacular definition to this excerpt of the patent’s abstract (emphasis ours):
[T]he social networking system may receive a list of proscribed content and block comments containing the proscribed content by reducing the distribution of those comments to other viewing users. However, the social networking system may display the blocked content to the commenting user such that the commenting user is not made aware that his or her comment was blocked, thereby providing fewer incentives to the commenting user to spam the page or attempt to circumvent the social networking system filters.
Amid the milieu of largely baseless U.S. congressional inquiries into perceived social media bias and a Republican Party frothing at any evidence, no matter how thin, that supports that crusade, Facebook laying legal claim to the act of automatic shadowbanning reads as a massive misstep.
As with other patents, this one attempts to cover as many related variations as possible, bestowing the power to shadowban on “a forum creator or moderator,” text or sentiment analysis software, or a “machine learning classifier.” Deployment examples include “profane words, extremely negative, racist or derogatory comments,” and as such it’s fair to assume an automated mass-hiding filter is not currently in use on Facebook.
Attempts have been made to clamp down on borderline content across Facebook’s apps, with Instagram most recently unveiling an AI-powered anti-bully initiative. The first element of this is a feature which appears to conduct sentiment analysis on potentially rude comments and attempts to sway users into not posting them.
A forthcoming feature more closely resembles the shadowbanning described in this patent, giving users the options to set other accounts as “restricted,” where restricted accounts will have comments they leave on the restricting user’s content hidden to everyone but themselves.
Jeff Kanter, the lead inventor on patent 10,356,024, worked as a product management lead at Instagram when the patent was first filed. (He has retained a similar title but since moved over to the Chan Zuckerberg Initiative — the philanthropic arm of Mark Zuckerberg’s empire.)
We’ve reached out to Facebook for comment and to learn what could possibly be gained by patenting a specific moderation tactic. We’ll update if we hear back.