SEMINAR DESCRIPTION & GOALS

Content moderation has become an increasingly controversial and necessary aspect 
of platform governance. Trust & Safety teams within the social media platforms 
have developed increasingly sophisticated technical and procedural mechanisms to 
enforce their platform’s community standards. Efforts to improve the enforcement 
of such standards, though, often requires the collection of more data from users. 
The trade-off between robust content moderation and respect for user privacy has 
received inadequate attention from platforms, policymakers, and legal scholars. 

This seminar provides students with the tools to assess the trade-offs between 
content moderation and user privacy. Students will gain enough familiarity with the 
technical aspects of content moderation to “speak engineer,” enough knowledge of 
content moderation adjudication systems to recommend reforms that respect user 
privacy while furthering the platform’s aims, and enough familiarity of applicable 
content moderation laws and policies to theorize and advocate for new regulations. 

Pre-Reading Materials

Required: Gillespie, Tarleton. 2018. Custodians of the Internet: Platforms, Content Moderation, and the Hidden Decisions That Shape Social Media. New Haven: Yale University Press.

Recommended: Suzor, Nicolas. 2019. Lawless: The Secret Rules That Govern Our Digital Lives. Cambridge: Cambridge University Press.

Other readings provided via PDF.

Assessment

Technical quiz and short answer exam (20%). Students will complete a multiple choice exam on the ins and outs of content moderation--specifically the technical aspects of content moderation. Students will also answer several short answer questions about the challenges of conducting content moderation at scale. 

Privacy Assessment of a Content Moderation Intervention (80%). Students will select a content moderation intervention from the Prosocial Design Network database and perform a privacy assessment on that intervention. The assessment must include the following aspects:

• A brief overview of the technical feasibility of the intervention
• An analysis of the privacy concerns raised by that intervention
• A series of recommendations for how best to mitigate those concerns
• A recommendation for whether platforms should adopt the intervention 

The assessment must exceed 2,500 words but may not exceed 3,500 words. We will discuss this assignment in more detail during the last class.