About the study
Federal and state election officials treat coordinated disinformation as a first-order threat to the 2026 midterms. The operations documented in the 2020 and 2024 cycles share a common structure: counterfeit official-results broadcasts monetized through donation scams, coordinated rumor cascades about counting and certification, synthetic impersonation of candidates and election officials, and synthetic narratives amplified by inauthentic engagement. They concentrate in the results period between election night and certification, when official outcomes are unresolved and counter-speech by campaigns and officials is at its weakest.
This study is the United States arm of the Electoral Information Integrity Monitoring program; the program's first study monitors Brazil's 2026 runoff (oam.berkeley.help). The goal is to close the measurement gap on the platforms where these operations are least visible to independent researchers. We combine lawful access to public content through the TikHub Academic API with transparent, auditable detectors adapted from Vigia, our prototype monitoring system built for the Brazilian election. The intended contribution is a validated detection method that civil-society monitors in other democracies can reuse.
Research questions
- Volume and timing. How does the volume and timing of election-related public content on TikTok, YouTube, X, and Instagram shift across election night and the results period, and which topics dominate each phase?
- Cross-surface diffusion. How quickly do documented operation classes (counterfeit official broadcasts, coordinated rumor cascades about counting and certification, synthetic-candidate impersonation, donation-scam funnels) move between the four platforms, and does diffusion accelerate while results remain unsettled?
- Detection validity. What precision and recall do rule-based and model-assisted detection methods achieve against researcher-verified ground truth for each operation class, and where do they fail?
Methods
The study runs in three phases: instrument development (through November 2, 2026), live measurement during election night and the results period (November 3–30, 2026), and validation and write-up (December 2026 through early 2027). All collection runs through the TikHub Academic API under the program's terms.
- Collection. Read-only TikHub Academic API queries in English (and Spanish where the API permits): keyword and hashtag panels (stop the count, rigged election, fake results, certification, candidate names, and swing-state county names), target panels of public accounts documented by US fact-checkers and election officials, and video and post search endpoints for each platform. TikHub's unified schema delivers every export as CSV, JSON, JSONL, or Parquet, and exact slices can be re-requested later so reviewers can reproduce published results. The study window runs September 1 through December 31, 2026, with September–November 2024 as a historical baseline covering the most recent general-election cycle.
- Detection. Transparent detectors: near-duplicate video hashing, near-duplicate text clustering, coordinated-timing cascades, impersonation heuristics, and audience-anomaly signals. Flags from our external livestream monitoring (Vigia, covering Twitch, Kick, and YouTube) are used as seeds to find cross-posted copies on the four TikHub platforms. No generative models are used, and every verdict is inspectable.
- Ground truth. Stratified samples of flagged clusters are independently coded against a published codebook. We report precision and recall per operation class, with detection latency for the live window.
- Analysis. Volume curves by class and window, network analysis of cross-account and cross-platform content reuse, monetization-path analysis of donation and payment links in counterfeit broadcasts, and detector evaluation.
Data protection & ethics
- Read-only charter. No posting, no account interaction, and no deanonymization of private individuals. Analysis is restricted to public content and platform-provided metadata.
- Program discipline. All queries execute under the TikHub Academic Research Program terms, scoped to this approved project. Records are held in access-controlled project storage, and only aggregate results leave the analysis environment. Raw content is not redistributed.
- Human subjects. The study analyzes public content at the aggregate level and does not profile individuals. IRB guidance is being sought through UC Berkeley's Office for the Protection of Human Subjects.
- Legal context. Research use complies with each platform's terms of service as accessed through TikHub's licensed API infrastructure, and with applicable US research-conduct requirements for analysis of publicly available data.
- Outputs. Open working papers, a methods paper on detector validation, and aggregated artifacts shared with US fact-checking and election-integrity organizations.
Team
Department of Industrial Engineering and Operations Research (IEOR), College of Engineering, University of California, Berkeley. His work applies operations-research and machine-learning methods to detection problems in large-scale information systems. He built Vigia, the prototype monitoring system this program extends.
Collaborators may be added through the TikHub academic program verification process.
Contact
ehovagim20@berkeley.edu
University of California, Berkeley
For correspondence regarding this research program.