Is AI Mission Critical?
A use-by-use analysis of oversight duties, with a case comparison, worksheet, and model board resolution.
The Center’s library
Research & Insights
Research papers, positions, submissions, essays, and reference tools on the decisions and accountability of AI oversight.
Core ideas state the Center’s governing positions. Papers develop the analysis and supporting evidence. Guides and tools help institutions apply the work.
The Authorities Register
Search laws, cases, regulations, and standards by industry, jurisdiction, and role, with source summaries and the Center's oversight interpretation.
The AI Oversight Governance Layer
The decisions that establish what an enterprise governs, who answers, and what record the program must produce.
Distributed Accountability
Why a shared public need requires specific responsibilities across institutions and named accountability within each.
Exhibit B Is the Oversight Exhibit
Recommendations and proposed additions to the AI Risk Evaluation Supplement, Version 5.0.
The Missing Oversight Layer in TACIP
The Center’s comment on the August 3 community discussion draft, with a separate summary sheet.
Cognitive Sovereignty and the Capacity to Decide
The human judgment an institution must retain to make an informed decision about AI.
Agentic AI and the Oversight Lens
The oversight implications of granting systems authority to act, and the limits of a model-only view.
Builders and Buyers
Decision rights and accountability for institutions that build AI and those that put it to work.
The Committee Fallacy
The anatomy of an oversight program and the responsibility of the committee that governs it.
The AI Regulatory Enforcement Landscape
The Center’s June 2026 analysis of regulatory expectations and enterprise oversight.
The Caremark AI Liability Roadmap
The Center’s earlier analysis of oversight duties. Read with the September research paper, Is AI Mission Critical?
The Two Duties
The Center’s analysis of the duties of care and loyalty in the AI environment.
A Tolerance Is Not a No
Risk tolerance as a decision point, and escalation that carries opportunity as well as threat.
The Capability to Take Risk
An oversight program as the institutional capability to take informed risks.
Governance Is the Accelerant
How a defined oversight program supports faster, informed decisions.
Deadlines move. The obligation does not.
The June issue on recording the basis for decisions made while regulatory deadlines are changing.
Assess the use. Record the judgment.
A single enterprise-wide label for AI leaves the board’s most important distinctions unresolved.
Accountability must survive the handoff.
A shared AI environment needs specific responsibilities and warnings that reach someone with authority to act.
The capacity to decide must be preserved.
An accountable person needs the knowledge and authority to question an AI-assisted recommendation.
The Informed Decision Standard
Can the organization demonstrate, in evidence, that it made an informed decision about AI risk?
Caremark Comes for AI
An early Center essay on oversight duties and AI. The September research paper develops the analysis use by use.
An AI Committee Is Not a Governance Program
Why forming a committee leaves the substance of its oversight still to be defined.
The Velocity Gap
The distance between the pace of AI adoption and the capacity to make accountable oversight decisions.
Who Decides, Who Answers
A RACI for decision rights and accountability across AI oversight.
AI Use Determination Worksheet
Appendix A of Is AI Mission Critical? The evidence to record for each AI use.
Model Board Resolution
Appendix B of Is AI Mission Critical? Illustrative language for recording oversight decisions.
AI Oversight Baseline Review
A reference for examining the institution’s oversight capability and the evidence supporting it.
Partner guideAI Oversight: Guidelines for Private Directors
Developed with the Private Directors Association.
Partner guideGuiding Principles for Programmatic AI Governance
Developed with the DCRO Institute.
Collaborative paperThe Resilience Imperative
Securing Communications Infrastructure as the Foundation of an AI-Powered Nation. A paper by Brian Allen and Mike Regan, published with TIA.
No publications match this search. Try a different term or select All work.