Hate speech has been a major problem on Facebook since its inception, with the issue intensifying after the election of President Donald Trump. In an effort to combat this rising trend, the social network has been instituting new policies and trying out new ways of identifying and preventing offensive content. Unfortunately, due to leaked documents, it’s been proven that these new measures are a joke.
The documents in question were published by ProPublica. Based on the content of the policy, it would seem that part of Facebook’s approach comes in the form of a quiz, whereby identifying certain characteristics of a group is needed in order to answer.
In one instance, slides presented “female drivers, black children and white men” with regards to the question of which group needs protection from hate speech. It would appear that the answer to this question is “white men.”
Danielle Citron, a professor of law then tells ProPublica that the quizzes and the policies, in general, seem to be a bit color-blind. That is, they don’t seem to make much of a distinction between oppressed groups and oppressive ones.
On that note, there is also the matter of contrary rules that Facebook has to contend with due to its global presence. As The Verge notes, the social media platform has no need to be bound to America’s First Amendment. It could censor as much as it wants, whenever it wants.
As of now, Facebook is walking a tightrope by maintaining a balance between a constructive tech company that promotes free speech while trying to remain as neutral as possible. After all, what’s hate speech in one country might not necessarily be the same in others.
More to the point, the company would have to take an ideological standpoint if it wants to define hate speech in any meaningful way. In the case of the “white man” answer, for example, the social network would have to figure out a way to differentiate offensive content aimed at specific individuals or an indistinguishable group. Otherwise, it would have to latch onto more generalized points of views.


Makemation: a Nollywood movie that shows AI in action in Africa
California's AI Executive Order Pushes Responsible Tech Use in State Contracts
NASA's Artemis II Mission: First Crewed Lunar Journey Since Apollo
Meta Ties Executive Pay to Aggressive Stock Price Targets in Major Retention Push
OpenAI Pulls the Plug on Sora, Ending $1 Billion Disney Partnership
Golden Dome Missile Defense: Anduril and Palantir Join Forces on Trump's $185B Space Shield
Federal Judge Blocks Pentagon's Blacklisting of AI Company Anthropic
Apple Turns 50: From Garage Startup to AI Crossroads
Elon Musk Announces Terafab: SpaceX and Tesla to Build Dual AI Chip Factories in Austin, Texas
Annie Altman Amends Sexual Abuse Lawsuit Against OpenAI CEO Sam Altman
Nintendo Switch 2 Production Cut as Holiday Sales Miss Targets
AWS Bahrain Region Disrupted by Drone Activity Amid Middle East Conflict
NVIDIA's Feynman AI Chip May Face Redesign Amid TSMC Capacity Crunch
Chinese Universities with PLA Ties Found Purchasing Restricted U.S. AI Chips Through Super Micro Servers
SpaceX Eyes Historic IPO at $1.75 Trillion Valuation
Palantir's Maven AI Earns Pentagon "Program of Record" Status, Reshaping Military AI Strategy 



