Four programs for research, measurement, education, and deployment
AIO operates through four coordinated programs. Research defines the core questions, development builds the measurement systems, education supports interpretation and use, and deployment carries findings into reports, RFCs, advisory work, and pilots.
PRISM and benchmark systems quantify how values, evidence, sources, and data selection appear in real model responses.
Reports, RFCs, training, and institutional collaborations move measured integrity profiles into public review and operating practice.
These programs are not isolated offerings. Together they form one operating loop from research to public application.
Concept papers and research notes define new integrity questions in Authority Stack terms that can be reviewed in public.
Benchmarks and PRISM analysis examine how real models prioritize values, evidence, sources, and data choices across actual scenarios.
Reports, training, RFCs, and partnerships carry measured findings into governance design, review procedure, and operating practice.
Four programs and their concrete destinations
We define AI integrity, the Authority Stack, authority pollution, and integrity hallucination as explicit concepts and testable hypotheses.
We build PRISM, benchmarks, audit workflows, and analysis interfaces for process verification.
We provide explainers and participation pathways so researchers, practitioners, and civic networks can interpret and use the findings.
We carry measured findings into reports, documents, partner workflows, and institutional pilots.