AI Security & Integrity Lab
Authorized defensive research on application security, software supply-chain risk, prompt-injection resilience, and information integrity.
View the security research program →Independent nonprofit research program
AI Understanding Research Group
We study how artificial-intelligence information is produced, sourced, organized, and understood. Our work combines computer science, public-interest data, source verification, and AI literacy.
Institution
AI Understanding
An independent U.S. 501(c)(3) nonprofit education and research organization. EIN 41-3273048.
Research lead
Research Program Lead
Founder of AI Understanding
Leadership and accountability
Principal-investigator equivalent
AI Understanding's founder leads this research group. As Research Program Lead, the founder serves as the principal-investigator equivalent for the independent group, accountable for its research agenda, project selection, methods, publication standards, open-data commitments, and corrections.
This is an organizational research role, not a claim of a university faculty appointment or an academic degree. AI Understanding is independent and is not a university.
Active work
Public research projects
AI News Observatory
A continuously updated study of AI news coverage, source domains, topic patterns, and the revision history of canonical stories.
View the observatory →Open AI Literacy Curriculum
A public framework that maps 30 AI-literacy competencies to courses, practice tasks, and applied capstones.
Explore the curriculum →Open research data
Record-level JSON and CSV releases with definitions, provenance, update dates, and public methodology for independent reuse.
Open the JSON dataset →Research practice
Methods and reproducibility
Traceable evidencePublished records retain canonical articles and primary evidence URLs.
Public definitionsMeasures, inclusion rules, limitations, and topic logic are documented.
Reusable outputsResearch data is released in stable machine-readable formats under CC BY 4.0.
CorrectionsMaterial changes follow a public corrections process and versioned methodology.