Research
Research questions, methods, contributions and evidence behind coauthored preprints.
AI in cybersecurity: the original evidence study
The research question, evidence matrix, contributions and limitations of our April 2026 preprint on AI in cybersecurity.
The Intervention Penalty: costs of AI agent checkpoints
A visual explanation of the Intervention Penalty's simulation hypothesis: checkpoint costs, criteria-based governance and the limits of the evidence.