Brian Peretti
Former CTO & Deputy Chief AI Officer, U.S. Treasury
An independent research & education institution
Defining oversight governance for AI in regulated industries.
The Center for AI Oversight equips boards, executives, regulators, and legislators to understand what must be governed, who is accountable, and what record supports an informed decision.
01What is governed.
02Who is accountable.
03On what record.
For boards and executivesFor regulators and legislatorsFor the professions that serve them
Why nowAI adoption is an enterprise decision.
AI adoption is an enterprise decision.
So is its oversight.
The Center's mission is to make accountability real wherever AI is put to work, so institutions can make informed decisions and answer for their consequences.
AI is entering consequential decisions faster than many institutions are defining the authority and accountability around it. Attention to the models alone leaves an enterprise question unanswered: who decided what the system may do, and who can change that decision?
Boards and executives need a documented basis for the risks they choose to take. The Center develops the ideas, research, and reference tools that make those oversight decisions explicit.
What the Center doesResearch that informs decisions.
Research that informs decisions.
Education that makes them possible.
Research & reference tools
Papers, legal and governance analysis, and practical reference materials for the decisions an institution must make about AI.
Explore the library →Standards & policy education
Documented contributions to public consultations and briefings on enterprise authority, accountability, and oversight.
Read the submissions →Education & convening
Board briefings, professional education, and conversations among directors, regulators, and practitioners.
Engage with the Center →The people informing the work
Meet the advisers →Our advisory board
David Simpson
Rear Admiral, U.S. Navy (retired); former FCC bureau chief
Chuck Senatore
Former Director, SEC Southeast Region
Jim Cunha
Former Executive Vice President, Federal Reserve Bank of Boston
Rudy Brioché
Former VP & Policy Counsel, Comcast; former FCC Chief of Staff and Legal Advisor
Three ideas at the center of the work
A clear basis for AI oversight.
01 / THE DECISIONS
The Oversight Layer
Technical controls govern the system. Management directs its use. Oversight governance establishes what must be governed, who answers for it, and the record the institution must be able to produce.
Read the Center's position →02 / THE RESPONSIBILITY
Distributed Accountability
AI operates through decisions made across many institutions. Accountability must reach each organization that builds, adapts, or puts it to work, and attach to named people within each. The public need is shared; each responsibility must be specific.
Why accountability must reach every institution →03 / THE RECORD
The Informed Decision Standard
The Center's test is whether the organization can demonstrate that it made an informed decision about AI. A traceable record allows an institution to take risk and explain the judgment behind it.
Read the essay →Selected work
View the full library →From the Center.
Research Paper · Legal & Governance Analysis
Is AI Mission Critical?
The Caremark question for directors and officers: which uses of AI create a duty of oversight, and what the record must show.
A use-by-use analysis, with a case comparison, determination worksheet, and model board resolution.
Open research resource
Explore the register →The Authorities Register
The laws, cases, regulations, and standards behind AI oversight. Search by industry, jurisdiction, and role, and distinguish source requirements from the Center’s interpretation.
Exhibit B Is the Oversight Exhibit
Proposed additions to the AI Risk Evaluation Supplement that identify the program, the accountable parties, and the decisions on record.
Read the submission →The Missing Oversight Layer in TACIP
The Center's proposal for a chartered oversight program, submitted in response to the August 3, 2026 community discussion draft.
Read the submission →In development · Q4 2026
About the auditing guide →AI Oversight Auditing Guide
A guide for auditors examining the design of an AI oversight program, its accountability, and the evidence that it operates in practice.
Research in collaboration
Partners & collaborations →Bringing oversight into the professions.
Publications with professional and industry associations connect the Center's research to the institutions that put AI to work.
The ideas in practice
The AI Oversight Program
The program is the governance. It records the scope the institution chose, the authority it granted, the people accountable, and the reporting and assurance it requires.
5Pillars
14Domains
76Controls
A structured reference for the decisions and evidence of oversight. The full Program sets out its domains, controls, and supporting sources.
The people behind the work
Brian J. Allen
Executive Director · Center for AI Oversight
An attorney and governance executive whose experience spans enterprise leadership, financial services policy, and the oversight responsibilities of boards. Brian founded the Center to define what AI requires of the institutions that put it to work.