Research

Research on AI, systems, and risk

Published reports and findings. Each one states what was examined, what the evidence supports, and what an operator should do with it.

Report

The Execution Frontier

AI catastrophic risk through 2030, examined through executable intelligence, delegated authority, control architecture and closed-loop systems.

Read the report →
Study

The Request-to-Operations Gap

How process, data, ownership, adoption and AI readiness connect behind a technology request, from two years of Begine Fusion delivery records.

Read the study →
Report

State of AI 2025

Where AI adoption, investment and regulation stood at the end of 2025, and what that means for business decisions through 2026.

Read the report →
Method

How We Measure AI-Assisted Work

Accepted output, review burden, rework, cycle time, risk and business value, measured against a recorded baseline.

See the method →
Coming soon

State of AI 2026

The next edition, rebuilt against the numbers that moved since the 2025 report.

Get it when it lands →