Detect Coordinated Inauthentic Behavior on a News-Sharing Network
Overview
What this challenge is about.
You receive a 60-day sample of about 6 million posts mentioning a recent election, with account metadata (creation date, posting times, follower graph). Design and prototype a CIB-detection pipeline combining temporal-pattern features (posting time entropy, burstiness), content-similarity features (near-duplicate detection), and graph features (co-mention community structure). Output a ranked list of suspicious clusters with evidence cards (which accounts, which posts, why). Validate against 5 known CIB clusters the platform's trust-and-safety team has already disclosed publicly. Deliver the pipeline, the evidence cards, and a 3-page methodology doc.
The Brief
What you'll do, and what you'll demonstrate.
Detect and document coordinated inauthentic behavior in election-related posts with evidence robust enough for publication.
Earning criteria — what you'll demonstrate
- Combine temporal, content, and graph features for CIB detection
- Design evidence cards that hold up to editorial and legal review
- Validate detection against publicly disclosed ground truth
- Document methodology for a non-technical defamation-lawyer audience
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
Data Scientist
Combining temporal, content, and graph features for a real investigative outcome is exactly the senior data-science work trust-and-safety teams hire for.
This challenge sharpens
- network-analysis
- anomaly-detection
- temporal-analysis
AI Safety Researcher
Documenting evidence chains for inauthentic-behavior detection sits at the intersection of trust-and-safety and AI safety research.
This challenge sharpens
- anomaly-detection
- evidence-design
- network-analysis
Applied AI Scientist
Translating detection methods into a defensible product (evidence cards) is the applied-AI scientist's daily work.
This challenge sharpens
- near-duplicate-detection
- evidence-design
- temporal-analysis