Flooding and Propaganda
Format. Break into groups of 4–5. Each group picks one of the two breakouts below (or takes both if time allows). Spend ~5 minutes skimming the prep reads, then ~10 minutes debating. A designated reporter brings the group’s position — and any dissents — back to the full class for a ~3-minute report-back.
Chapter 3.1–3.3.6 argues that modern information control is often not about removing content but about drowning it out and amplifying selected narratives. The two motions below force groups to draw lines that the book deliberately leaves unresolved: where does organic speech end and coordinated propaganda begin, and what should platforms do when the speaker is a state?
Breakout A: The Line Between Amplification and Astroturf
Motion. “On today’s algorithmic platforms, ‘organic virality’ and ‘coordinated propaganda’ are indistinguishable in practice — and treating them as separate categories is a policy fiction.”
Prep reads (5–10 min).
- Meta Quarterly Adversarial Threat Report, Q1 2025 — Meta Transparency Center, 2025. Meta’s own “kill chain” framing of CIB takedowns, with case studies from Iran, Russia, and cross-platform migration.
- How Coordinated Inauthentic Behavior continues on Social Platforms — Stanford FSI, 2025. Stanford researchers document that Russian Doppelganger accounts flagged by Meta remained active on X and TikTok, suggesting cross-platform threat-sharing has collapsed.
- Coordinated Inauthentic Behavior on TikTok: Challenges and Opportunities for Detection in a Video-First Ecosystem — arXiv (ICWSM), 2025. Analyzes 2024 election-cycle CIB on TikTok — synchronized amplification, AI voiceovers, manufactured split-screens — and how behavioral signatures blur into engagement-farming.
- Polish presidential election Doppelganger targeting report — EU DisinfoLab, April 2025. 279 Doppelganger posts targeting Poland’s May 2025 election with anti-EU / anti-Ukrainian framing; shows the tactic set persisting despite years of takedowns.
Discussion prompts.
- The book identifies four behavioral signatures of propaganda campaigns: high volume, high retweet ratio, fast retweets, coordinated timing. Modern engagement-farming influencers exhibit all four. Does that mean they are propaganda, or that the signatures no longer discriminate?
- China’s 50 Cent Party and Russia’s IRA both employed real humans posting real opinions for money. If the posts are sincere but paid, is that flooding? Astroturfing? Just marketing?
- Benkler’s 2016 finding was that bot amplification did not meaningfully shift users’ source diet — mainstream media still dominated. Does that finding still hold in a TikTok-dominated 2026 information environment? What would it take to redo that study today?
- If a platform cannot reliably tell a coordinated campaign from a viral trend, should the default be permissive (let it run) or restrictive (throttle unfamiliar spikes)? Who wins under each default?
Bring back. Your group’s best operational definition of “coordinated inauthentic behavior” that a platform could actually enforce — and the biggest false-positive case it produces.
Motion. “Platforms should label all state-affiliated media accounts (RT, CGTN, VOA, BBC, Al Jazeera, PressTV, etc.) with a uniform disclosure — and neither ban them nor leave them unlabeled.”
Prep reads (5–10 min).
- Holding the line: Auditing the EU’s ban of Russian state media 3 years on — Institute for Strategic Dialogue, 2025. Documents that RT, Sputnik, and 30+ sanctioned outlets remain widely accessible across EU member states during the Feb 2025 German and May 2025 Polish elections.
- Catch Me If EU Can: How RT and Sputnik Evade EU Content Bans — Alliance for Securing Democracy / GMF, 2025. Maps 308 mirror sites, proxy X/Telegram accounts, and Apple/Spotify podcast feeds that keep RT and Sputnik content flowing to EU audiences.
- A 65-language megaphone: how Chinese state-media CGTN is shaping global narratives — Reporters Without Borders, 2025. Describes CGTN’s cross-platform reach into 160+ countries and inconsistent labeling on Facebook, Instagram, and YouTube.
- How Meta Platforms Ireland may have channeled money to RT and Sputnik via revenue-sharing — Whattofix, June 2025. Investigation finding RT Arabic and Sputnik pages remained active in Meta’s revenue-share programs long after EU sanctions — a labeling-vs-sanctions collision.
Discussion prompts.
- The chapter opens with Russian Telegram channels spreading a fabricated “Ukrainian soldiers burning Russian books” video during the 2022 invasion. Would a “state-affiliated” label on the originating account have changed anything about how it spread through pro-Russian networks?
- What counts as “state-affiliated”? BBC receives a UK license fee. NPR receives some federal money. Does the label mean the same thing applied to CGTN, RT, VOA, and BBC? Should it?
- Elon Musk’s X removed state-media labels in 2023, arguing they were being applied inconsistently. Was that a principled position or a policy retreat? What would consistent enforcement even look like?
- Compare three regimes: (a) EU-style forced labeling under the DSA, (b) US-style laissez-faire with occasional deplatforming, (c) the Brazil model where a court orders a platform-wide block. Which produces the most legitimacy? The least collateral damage?
Bring back. A one-sentence policy your group would ship if you ran trust & safety at a major platform, and the strongest objection to it.
Instructor notes
These map to two of the “Section takeaways” boxes: the flooding section’s argument that flooding is invisible to legal frameworks built around removal, and the propaganda section’s finding that behavioral signatures of influence campaigns predate their political salience. Breakout A tends to surface the definitional problem (students often start certain there’s a clear line, then lose it); Breakout B tends to surface the double-standards problem (students who defend labeling RT often struggle with BBC). If time is short, A is the deeper technical debate; B is the more concrete policy debate.