Financial Integrity Watch reconstructs consequential AI and cyber events as chains of action: what reached which system, under what authority, how the system responded, what evidence survived, and whether money, records or recovery were affected.
What we publish
Dispatches explain current incidents. Updates carry only material changes. Explainers teach durable concepts once. Pattern Reviews compare incidents only when the comparison creates new knowledge. Governance Briefs cover decisions that change authority, access or institutional expectations.
How we report
Research rigor earns our confidence internally. Editorial clarity earns the reader’s attention externally. We distinguish controlled tests from production incidents, attempted actions from completed results, and model assistance from autonomous action.