Algorithm Watch
The systems that decide. Watched.
Algorithm Watch is our investigative laboratory for the algorithmic era — auditing AI systems, recommendation engines, moderation policies, and automated public-sector decisions for bias, error, and harm.
Active Investigations
Currently under review.
AW-2026-014
Recommendation Suppression on Major Video Platform
Platform
Active
AW-2026-011
Hiring Algorithm Disparity at Three Fortune 100 Employers
Employment
Active
AW-2026-009
Facial Recognition Misidentification — Municipal Audit
Public Sector
Reporting
AW-2025-041
Generative Image Model Stereotype Benchmarks
AI Models
Published
AW-2025-036
Loan Underwriting Model Audit
Finance
Published
Lab Modules
Seven ongoing lines of inquiry.
01
AI Bias Reports
Empirical testing of frontier and production AI systems.
02
Platform Transparency
Annual scorecards on moderation, appeals, and disclosure.
03
Creator Economy Analysis
Distribution, demonetization, and discoverability outcomes.
04
Recommendation Studies
Long-running observation of feed dynamics.
05
Facial Recognition Audits
Independent benchmarking in public deployments.
06
Hiring Algorithm Analysis
Resume screeners, video interviews, scoring tools.