Explaining Twitter's inability to effectively moderate content during the COVID-19 pandemic

Area of Science:

• Social Media Studies
• Public Health
• Information Science

Background:

• Social media platforms face challenges balancing content moderation with free speech.
• Platform design may impact the effectiveness of content moderation strategies.
• During the COVID-19 pandemic, platforms removed medical misinformation, but efficacy remains unevaluated.

Purpose of the Study:

• To evaluate the efficacy of Twitter's content moderation policies in curbing vaccine misinformation during the COVID-19 pandemic.

Main Methods:

• Analysis of vaccine-skeptical and pro-vaccine Twitter account clusters.
• Assessment of tweet volume, virality, and content quality before and after moderation actions.
• Examination of a large-scale account removal event on January 8, 2021.

Main Results:

• Vaccine-skeptical clusters generated more vaccine-related tweets and spread lower-quality content despite moderation.
• Sustained content and account removals did not significantly reduce misinformation spread from skeptical clusters.
• A large-scale account removal event temporarily reduced targeted activity but increased virality and decreased information quality among related accounts.

Conclusions:

• Twitter's interventions, including deplatforming and content removal, showed limited efficacy in curtailing vaccine misinformation.
• Content moderation strategies may inadvertently amplify misinformation or shift it to related accounts.
• Platforms with similar architectures may encounter comparable challenges in moderating harmful content effectively.