AI Analysis of Journalism's Network Structure
From the COVID-19 Crisis to Today
During the first months of the COVID-19 pandemic in 2020, news organizations occupied an unexpected position in digital information networks. Our team's analysis of 141 million tweets from 24 million users revealed that journalism functioned as connective infrastructure on Twitter, bridging otherwise disconnected communities across ideological, geographic, and linguistic divides. News accounts created the structural links that held fragmented publics together during a period of profound uncertainty.
By 2023, the landscape had shifted. In qualitative interviews with 45 younger news consumers across the United States, Nigeria, and India, we found that while the pandemic initially drove audiences toward trusted sources, Gen Z had developed new strategies for navigating information. They cross-checked claims across multiple platforms, treated news as one layer in distributed sensemaking rather than a singular authority, and increasingly turned to TikTok, Instagram, and YouTube creators for context and analysis. Trust in legacy institutions had declined, and news consumption had become more fragmented and socially mediated.
Now, in 2025, we have collected new survey data from hundreds of respondents across these same three countries, documenting current news consumption patterns, platform preferences, and trust dynamics. Our three layers of data from 2020, 2023, and 2025 will allow us to conduct the first longitudinal study tracking journalism’s network role and audience relationships across five years of dramatic social and technological change, from pandemic crisis through post-pandemic adjustment to today’s AI-saturated media environment.
This Knight Lab studio project will analyze these three datasets together to tell one comprehensive story about the news on digital platforms today, examining whether journalism maintained its bridging function, how content strategies evolved, and what these patterns reveal about news in networked publics. Our prior research has been supported by the Mellon Foundation and Google News Initiative and has produced peer-reviewed publications on network architecture and platform journalism. Students in this studio will be co-authors on all academic articles published from this work, gaining direct experience with the full research publication process, and adding unique research products to their resumes.
Faculty and Staff Leads
Project Details
2026 Winter
Important Questions
- How has journalism's structural position in social media networks changed from 2020 through 2023 to 2025, and what explains these shifts?
- Which platform-native content strategies have persisted or emerged across this period, and how do they correlate with journalism's network centrality?
- Do Gen Z news consumption patterns reported in surveys align with observable network structures, and where do self-reported behaviors diverge from actual information flows?
- What does longitudinal analysis reveal about journalism's long-term viability as digital infrastructure in an era of fragmented platforms, AI-generated content, and declining institutional trust?
Sample Milestones
- Weeks 1-2: Explore existing datasets from 2020 Twitter analysis, 2023 qualitative interviews, and 2025 surveys. Conduct focused literature review on longitudinal media studies and network evolution. Develop analytical frameworks for comparing findings across time periods and methods.
- Weeks 3-5: Learn network analysis tools and comparative methods. Begin integrated analysis across all three datasets, identifying patterns in how journalism's network position, audience behaviors, and content strategies have evolved. Work in teams to map structural changes and behavioral shifts.
- Weeks 6-8: Conduct deep comparative analysis connecting network structures with audience practices. Synthesize findings across quantitative and qualitative data to develop arguments about journalism's transformation.
- Weeks 9-10: Finalize manuscripts for journal submission, create data visualizations, and prepare public-facing research summaries. Present findings at Knight Lab showcase. Submit co-authored work to target journals.
Outcome
This studio project will produce the first longitudinal study examining journalism's evolving role in digital networks across five years. Students will serve as co-authors on academic manuscripts while developing skills in AI-driven computational analysis, mixed-methods research, and publication-quality research methods.