Facebook: Leveraging AI to address Content Moderation and Data Privacy Challenges
| Case Code: ITSY144 Case Length: 14 Pages Period: 2024 Pub Date: 2024 Teaching Note: Available |
Price: Rs.400 Organization: Facebook Industry: Technology & Communications Countries: United States Themes: Artificial Intelligence, Information Systems, Collaboration Tools |
Abstract Case Intro 1 Case Intro 2 Excerpts
Excerpts
AI Tools
Facebook implemented a number of products to combat misinformation and ensure user safety and for recommendation of content and predictive analytics through the use of AI. The company’s AI teams built machine learning models to analyze photos, texts, and videos in different languages as well. In short, AI played a crucial role in the content review process. For instance, in 2013, Facebook introduced EdgeRank as a method to rank and display content in the users’ News Feeds. This was a relatively simple algorithm and considered 3 factors namely: Affinity (the user’s relation with the content creator), weight..
New Language
In June 2017, researchers from Facebook Artificial Intelligence Research (FAIR) published research on the introduction of dialog agents that had the capability to negotiate. In this article the researchers discussed the challenges of communication and reasoning required for negotiations . These researchers developed a method using reinforcement learning and conversation simulations to train chatbots in negotiations. These chatbots were exposed to a massive dataset of negotiations done by humans that included the words spoken and the final decisions..
Misinformation And Fake News
The Covid-19 virus originated in China in December 2019, and spread across many provinces in the country and gradually across the globe. The virus affected people globally, leading to millions of deaths. Many governments implemented lockdowns and social distancing to curb the spread of the virus. During this time, social media was flooded with millions of posts about the Covid-19 virus, vaccines, and deaths..
Algorithms Fueled Gender Discrimination
In 2021, research by the University of Southern California revealed that Facebook’s ad delivery system had a gender bias. The research team investigated advertisements from software engineering to sales to food delivery. To understand whether the ad delivery was skewed due to gender bias the researchers compared two similar job ads. First, both ads were for positions with the same skillsets. Second, the ads were for a group of companies that had a predominantly male workforce and for another group with a predominantly female workforce..
Fair Housing Act
In June 2022, the US Department of Justice announced a settlement agreement with Facebook to address allegations made by the Fair Housing Administration (FHA) . The lawsuit alleged that Facebook’s housing advertising system discriminated against users based on race, color, religion, sex, disability, familial status, and national origin..
Looking Ahead
Facebook’s multi-faceted approach to tackling misinformation included advanced algorithms, fact checking, user reports, and moderation of content. This led to a noticeable reduction in the spread of fake news. Yet, the system was not totally effective in controlling news. Some industry analysts wondered whether Facebook’s system architecture was to be blame, as the social media platform unintentionally encouraged users to like, comment on and share content that contributed to the issue..
Exhibits
Exhibit I: A Brief Snapshot of the Community Standards Enforcement Report
Exhibit II: Second Quarter 2024 Financial Highlights
Exhibit III: Facebook’s Community Standards
Exhibit IV: Example of the bot Negotiations that Meta Researchers Observed
Exhibit V: Details of a few of the AI tools used by Facebook
Exhibit VI: Ad Runs for Delivery Drivers
Buy this case study (Please select any one of the payment options)
| Price: Rs.400 |
Price: Rs.400 | PayPal (9 USD) |
