Who governs AI
AI governance isn't one office. It's a standing table of stakeholders, each seeing the same initiative from a different seat. Here's who sits there, what each owns, and a decision-rights matrix so every consequential call has one accountable owner and the right voices consulted before, not after.
The seats at the table
Each seat has a deep page, the same framework, seen from that role's vantage.
Cabinet & board
Strategy, principles, policy adoption, and funding the program.
OpenAI & governance leads
Convene the table, prioritize the portfolio, and right-size review.
OpenFaculty & academic governance
Classroom use, academic integrity, and academic AI policy.
OpenResearch & sponsored programs
Research integrity, human-subjects review, and funder/export compliance.
OpenInstitutional research & analytics
The predictive models, retention, risk, enrollment. That act on students.
OpenIT, data & security
Tools, data classification, security, deployment, and incidents.
OpenLegal & compliance
Legal applicability, contracts, records, and defensibility.
OpenFinance & budget
Funding model, the business case, and financial terms.
OpenHR & people
Workforce-facing AI, hiring tools, and the duty to bargain.
OpenStudent affairs & enrollment
Student-facing AI, chatbots, advising nudges, admissions screening.
OpenLibrary & information services
AI literacy, source evaluation, licensing terms, and reader privacy.
OpenStudents
A real voice in student-facing decisions; access and equity.
OpenUnit leads & practitioners
Bring the use case and run it through the framework.
OpenDecision rights, who does what
For each key decision, who's Responsible, Accountable, Consulted, and Informed. The point isn't the exact letters. It's that no decision is orphaned, and none is made in one office alone.
| Decision | Cabinet | AI lead | Faculty | Research | IR / Data | IT / Sec | Legal | Finance | Student Aff. | HR |
|---|---|---|---|---|---|---|---|---|---|---|
| Set AI strategy, principles & policy | A | R | C | C | C | C | C | C | C | C |
| Adopt or buy an AI tool | I | A | C | · | · | R | C | C | · | · |
| Approve a high-risk system | I | A | C | C | C | R | C | C | C | C |
| Classify data & run security review | · | I | · | · | C | A | C | · | · | · |
| Set classroom & academic-integrity use | I | C | A | · | · | · | C | · | I | · |
| Use AI in research or with human subjects | · | I | C | A | C | C | C | · | · | · |
| Run a predictive model on students | I | C | C | · | A | C | C | · | R | · |
| Deploy student-facing AI (advising, admissions) | I | C | · | · | C | R | C | · | A | · |
| Use AI in hiring or HR decisions | I | C | C | · | · | C | C | · | · | A |
| Fund it (budget & sustainability) | A | C | · | · | · | I | · | R | · | · |
| Respond to an AI incident | I | C | · | · | · | A | C | I | C | C |
| Communicate & disclose to the community | A | R | C | · | · | I | C | · | C | · |
Students are consulted on any student-facing decision and informed throughout; unit leads and practitioners bring the use cases that enter this table. Accessibility (ADA/504), privacy, and equity are cross-cutting requirements on every row, not a single seat, build them into each decision. Adapt the seats and assignments to your own titles and authority, the point is that every consequential AI decision has one accountable owner and the right voices consulted before, not after.
See the full Roles & Responsibilities pillar