Facebook and disinformation

Facebook (now Meta Platforms) is the world's largest social network, with 3+ billion monthly active users. As a primary news source for many users and a central hub for political discourse, Facebook has become a major vector for disinformation, coordinated manipulation, and false information spread. The platform's algorithmic amplification, group structures, and network effects create affordances for both organic viral spread and deliberate coordinated campaigns.

Platform mechanics enabling disinformation

Algorithmic feed: The feed prioritizes engagement (likes, shares, comments) over accuracy, creating incentives to post sensational or emotionally provocative content—including false claims.

Groups and pages: Facebook's structure of pages (one-to-many broadcast) and groups (community discussion) enables both organic communities and coordinated networks to form and amplify messaging.

News Feed amplification: Shared URLs and content receive algorithmic boost; coordinated simultaneous sharing can exploit trending signals and reach broader audiences.

CrowdTangle data: Until 2024, Facebook provided the CrowdTangle API, allowing researchers to track public shares and network activity; now deprecated, it was the primary tool for studying coordinated behavior on the platform.

Limited transparency: Despite public commitments to reducing disinformation, Facebook limits third-party access to data needed for detection and intervention.

Key papers

• Sahly et al. (2019) — Social Media for Political Campaigns: An Examination of Trump's and Clinton's Frame Building and Its Effect on Audience Engagement — Comparative analysis of Trump's (303 posts) and Clinton's (352 posts) Facebook content during 2016 campaign; examines frame effects on engagement (shares, comments, likes/reactions); finds platform-specific engagement patterns compared to Twitter

Challenges

Scale: Billions of daily posts make comprehensive content moderation infeasible; platform relies heavily on user reports and automated flagging.

Coordinated networks: Accounts and pages can be created cheaply and in bulk; networks coordinate across regions and languages; deception through entertainment and local-interest framing evades detection.

Political polarization: Facebook's algorithmic feed and group structures can amplify polarized, partisan content; coordinated campaigns exploit existing divisions.

Jurisdictional issues: Facebook operates globally but faces different regulations (EU GDPR, US Section 230, national election laws) that complicate enforcement and researcher access.

Key papers in this wiki

• Cinelli et al. (2021) — The echo chamber effect on social media — Demonstrates that Facebook's feed algorithm drives homophilic clustering and polarized information diffusion more strongly than community-based platforms, with users receiving information primarily from like-minded peers
• Bode & Vraga (2015) — In Related News, That Was Wrong: The Correction of Misinformation Through Related Stories Functionality in Social Media — tests "related stories" algorithm as a corrective mechanism; shows related stories reduce health misinformation misperceptions when matched to user's initial position, suggesting algorithms can contribute to correction beyond content moderation
• Giglietto et al. (2020) — It takes a village to manipulate the media — detects coordinated inauthentic behavior on Facebook during Italian elections via near-simultaneous link sharing; shows coordinated networks amplify problematic content 1.79–2.22× more than uncoordinated entities

Related topics

• Media manipulation — broader coordinated campaigns beyond Facebook
• Coordinated inauthentic behavior — coordinated networks masquerading as authentic users
• Social media and misinformation — role of platform affordances in disinformation spread
• Election interference — coordinated campaigns targeting political outcomes

Open questions

• How have changes to Facebook's transparency policies (deprecation of CrowdTangle, data restrictions) affected research capacity to detect coordinated disinformation?
• What fraction of disinformation on Facebook originates from coordinated campaigns vs. organic user creation and spread?
• How effective are Facebook's current interventions (labels, account suspension, group moderation) at reducing coordinated disinformation without suppressing legitimate coordinated activism?