Data & Society presents Robyn Caplan’s map of three content moderation approaches

Data & Society presents Robyn Caplan’s map of three content moderation approaches

Robyn Caplan’s report describes three ways platforms organize content moderation: artisanal teams that work manually in-house, community-reliant platforms that use volunteers to enforce policies, and industrial operations that use automated technologies and separate policy development from enforcement. The categories connect how moderation is organized to a central tension: responding to speech in context while applying rules consistently at scale. Caplan argues that platforms’ missions, business models, team sizes, and organizational dynamics shape their approaches. Each approach carries different trade-offs and resource needs, so the report says public oversight should consider both the context of speech and how platforms are organized—not just the rules for particular types of content. The report draws on interviews with major digital platforms, but this page does not detail the interview methods or show how the approaches affect moderation outcomes. What organizational conditions help platforms balance context-sensitive decisions with consistency at scale?

Publication: 236f5fc6-049c-4fcd-8507-db38110e49d3

attribution — pass: The draft attributes the report, its three-part typology and the oversight argument to Robyn Caplan, and names Data & Society as publisher. This matches the page. Claims about missions, business models and team sizes are framed as the report's analysis, which the page supports.

“Public oversight,” writes Caplan, “needs to incorporate not only the context of speech, but the organizational dynamics of platforms

summary — pass: The draft accurately summarizes the artisanal, community-reliant and industrial approaches and the context-versus-consistency trade-off. It is readable without specialist background. The phrase 'not just the rules for particular types of content' slightly reframes the source, which links organizational dynamics to deciding where new rules and resources are needed. This is not a contradiction. The question follows from the evidence, and the limitations accurately note what the page lacks.

each approach has different trade-offs between moderating content in context and maintaining consistency across a large scale

date — pass: The originalPublished value of 2018-11-14 matches the page's stated publication date. The draft makes no conflicting date claim.

Published on November 14, 2018

support — pass: Each claim has a matching verbatim excerpt. These include the three strategy types, manual in-house work, the trade-offs, the oversight argument and interviews with 10 platforms. The draft invents no benefits, costs or outcome claims, and it explicitly says outcomes are not shown.

drawing from interviews with 10 major digital platforms.

media — pass: The selected media is the verified direct link to the original Data & Society report page. This is the most direct way for readers to reach the source.