For campus leaders: presidents, provosts, CIOs, deans, and boards
The accountability for AI is already yours. The work can be delegated; the accountability cannot.
It reaches into decisions already yours: what your degree certifies, how you steward institutional data, who keeps their job, where the money goes, and your reputation when something goes wrong.
None of that is new. AI only raised the stakes. Handling it well takes capability at three levels that build on each other, Literacy, Competency, and Fluency, and the gap between them is where institutions get it wrong. This site helps you close it.
New to the framework? Start at campusaiproficiency.com for the full AI Proficiency Continuum. This site is its leadership path, the Fluency layer read for those who set direction.
Your working guide to leading AI: a framework, ten questions for your next cabinet agenda, governance instruments you can adopt, and resources your team can use on its own.
You, if you are accountable for AI decisions you cannot delegate: president, provost, CIO, dean, governance officer, cabinet member, or board, at any institution or system.
So you can take a defensible position, name an owner, and set a date, instead of deferring to whoever sounds most technical in the room.
Literacy: your floor for credible oversight
You cannot govern what you cannot describe. You need to be able to say what these tools do: produce plausible output rather than verified fact, inherit the bias in their training data, and make the data terms in a contract its most consequential clause. This is not a technical course; it is the working knowledge you already carry on finance or enrollment. Compact, and here in full.
Competency: how you use it in your own work
Your office handles the institution's most sensitive matters, so it carries the tightest limits on what these tools may touch. Used with discipline, AI earns its place: it synthesizes at volume and stress-tests a decision before you make it. Used without discipline, your office becomes the institution's most exposed and most visible point of risk. Your people calibrate their conduct to what they watch you do.
Fluency: the decisions only you can make
Who holds which decision rights. The standards you set before the first vendor call. How your data may be used, decided as mission rather than technical default. The nerve to fund what serves you and end what does not. Real sponsorship: your name on it, defending the budget, taking the call when something breaks. Many can advise you here. Only you are accountable, and this is the level the site serves.
A note on the word. Institutions use "fluency" differently; some apply it to graduates, some to the whole community. Here it carries the meaning defined in the AI Proficiency Continuum: the strategic tier, the adaptive expertise to evaluate, adopt, and govern AI at institutional scale.
Why a generic playbook will not work for you
You do not run a typical enterprise, so you cannot lead AI as if you did. Your authority is genuinely shared: faculty hold curricular ground, your senate holds real voice, and academic freedom is a feature you protect. Leading AI well means knowing which decisions belong to shared governance and working through those bodies. Hand a direction down from the top and it will be resented, and it will not stick. This site shows you how to exercise that judgment.
Written for the governance realities of community colleges, comprehensive and research universities, independent institutions, and systems.
Start · The Framework
One Continuum, Read at Leadership Depth
This site is the leadership expression of the AI Proficiency Continuum, the capability framework published at CampusAIProficiency.com and part of the Campus AI Framework.
The continuum defines three cumulative layers of AI capability: Literacy (the shared foundation, for everyone), Competency (applied capability, by role), and Fluency (strategic capability, for those who set direction). These run through six domains: AI Foundations, Responsible Use, Applied AI, Data Literacy for AI, AI Governance & Policy, and Strategic AI Leadership.
This site expands three rows of that matrix for one audience, senior institutional leaders and sponsors:
The layers are cumulative. Competency assumes Literacy; Fluency assumes both. Leaders don't get to skip the base. What they get to skip is the semester.
Licensing
The framework and everything on this site are published under CC BY-NC-SA 4.0, the same license as the parent framework. Take it, adapt it, run it on your campus, with attribution.
The Three Layers · Layer One
The Foundation: Literacy, for Leaders
Fluency is built on literacy. The layers are cumulative, and leaders don't get to skip the base. What you get to skip is the semester.
Below is the whole foundation, all six domains, at the depth your role requires: enough to use AI safely, question it well, and never nod at a term you don't understand. Read it in under an hour. Everything deeper on this site assumes it.
Domain 1
AI Foundations: What these systems are
AI systems predict; they don't understand. A large language model produces the most plausible next words, usually right, sometimes confidently wrong, never self-checking. Capability is jagged: brilliant at synthesis, capable of failing arithmetic or inventing a citation, and the boundary isn't intuitive. Agentic systems raise the stakes: they don't just answer, they act. The essential vocabulary: model, prompt, hallucination, fine-tuning, RAG, guardrails, agents, context window.
You're literate here when you can explain to your board, in two sentences, what AI is good and bad at, and why it must be verified.
Domain 2
Responsible Use: The lines that don't move
Bias is an inheritance from training data, not a malfunction. Privacy obligations (FERPA, HIPAA where applicable) mean student and health data never enters unapproved tools, and consumer versions of AI tools are almost never compliant. Academic and research integrity require disclosure and attribution. Accessibility applies to AI-generated content too. These aren't aspirations; several are law.
You're literate here when you can name the three kinds of data that must never touch an unvetted tool, and you'd catch the risk in "staff are just using the free version."
Domain 3
Applied AI: Using it yourself
You cannot govern what you've never touched. Basic applied literacy for a leader: write a prompt, iterate it, and verify the output before relying on it, on real work, weekly. Drafting, summarizing, and preparing for meetings are the natural executive entry points. Imperfect, occasionally embarrassing use is the point; secondhand understanding produces secondhand judgment.
You're literate here when you used an approved AI tool this week on real work, and caught at least one thing it got wrong.
Domain 4
Data Literacy for AI: What feeds the machine
AI is only as good as the data underneath it, and most institutional data is messier than its owners believe. Know your institution's data classification tiers (what's public, internal, restricted) because they determine which tools may touch what. Read AI-generated analyses critically: a confident chart built on bad data is worse than no chart.
You're literate here when you ask "where did this data come from and would we make a decision on it?" before accepting any AI-produced analysis.
Domain 5
AI Governance & Policy: The rules of your own house
Know what your institution's AI guidance actually says: what's allowed, what's approved, where the lines are, and where a colleague would find it. Know that "human in the loop" is meaningless until someone specifies which human, with what authority and time. Know that regulation is arriving: some jurisdictions now mandate AI literacy itself.
You're literate here when you can state your institution's current AI policy in one sentence, or state honestly that it doesn't exist yet, which is its own finding.
Domain 6
Strategic AI Leadership: Why this reaches your desk
AI is reshaping the sector's economics, workforce, and expectations. 86% of higher-ed staff already use it, mostly untrained, and your peers are making adoption moves that will bear on your enrollment and reputation. Literacy at this domain means seeing AI as an institutional force, not an IT topic.
You're literate here when you can say what AI has to do with your strategic plan, in mission terms, unprompted.
Reference Appendix
Conversant, Not Technical
The page to read the night before an AI agenda item: five ideas that explain almost everything, the vocabulary as you'll hear it in meetings, what these systems cannot do, and the commitment to staying current.
Open the reference →The Foundation · Reference
Conversant, Not Technical
The page to read the night before an AI agenda item.
Part 1: Five ideas that explain almost everything
A large language model produces the most plausible next words, usually right, sometimes confidently wrong, never checked against truth unless a human or another system does the checking. What it changes for you: any use where a wrong-but-confident answer causes harm needs human review by design, not by policy memo.
Fed the internet, it inherits the internet's biases; fed historical institutional data, it inherits historical patterns, including the ones you've spent a decade correcting. Bias isn't a malfunction; it's an inheritance. What it changes for you: "has this been tested on a population like ours?" is a fair, answerable question.
Simultaneously better than experts at some tasks and worse than a first-year student at others, and the boundary isn't intuitive. What it changes for you: every proposed use case needs its own evidence. "It's good at X" says nothing about Y.
The most consequential questions in any purchase are about data: what goes in, where it's stored, whether the vendor trains on it, what happens at contract end. What it changes for you: data terms are a leadership review item, not a procurement checkbox.
Model capabilities, prices, and vendor lineups change every few months, which is why rigid multi-year plans age badly. What it changes for you: build revision into the calendar, and never let a vendor's current demo anchor a five-year commitment.
Part 2: The vocabulary you'll actually hear
For each: what it means in plain terms, and the question to ask when you hear it.
A general-purpose text-prediction engine products are built on. Ask: which model, whose, and what happens to us when they change or retire it?
They further trained a model to specialize it, sometimes just marketing. Ask: on what data, with whose permission, and how do you measure it's better for our population?
It retrieves from a defined document set and answers from that. Ask: what's in the set, who maintains it, and what does it do when the answer isn't there?
A fluent, confident, false output, inherent to prediction, reducible but not eliminable. Ask: your measured error rate on tasks like ours, and the review process for what gets through?
Filters bolted around the model, useful, imperfect, routinely bypassed. Ask: against what, tested how, and the escalation path when they fail?
It takes actions instead of only answering, a categorical jump in risk because errors become actions. Ask: what can it do without a human approving, and how do we undo it?
The single most important contract sentence. Ask: show me that clause, and show me it applies to every tier of the product we're buying.
Meaningless until specified. Ask: which human, with what training, what authority to override, and what time to actually do the review?
Part 3: What these systems cannot do
AI produces outputs; only people bear responsibility. Every "the algorithm decided" is an accountability gap someone chose to create.
It has no idea what your mission, politics, history, or students require unless people encode that, which is why staff's tacit knowledge becomes more valuable, not less.
No system reliably knows when it's wrong. Verification is a human function you must staff and design for.
Choosing, with values, tradeoffs, and consequences, remains the job you were hired for. That job is not being automated; it's being raised in stakes.
Part 4: The commitment: staying current in your scope
Being conversant is a standing commitment of the role: you are responsible for knowing what's changing in AI within your areas of accountability, and for knowing where reliable information about it lives. Not daily news-chasing, a deliberate rhythm built from named sources.
Know your sources, by scope. Sector-wide: EDUCAUSE, the higher-ed press, peer networks. Governance: AGB, your counsel's regulatory updates. Policy: NIST's AI work and your state's legislative activity. Portfolio-specific: the association serving your function. The failure mode isn't reading the wrong things, it's having no named sources at all.
Build the rhythm. A recurring quarterly block to take stock; one standing leadership-agenda item where your people surface what's moving; and your own hands on the tools weekly, because thirty minutes of firsthand use beats thirty articles.
The leaders who stay conversant aren't the ones who read the most; they're the ones with named sources and a rhythm.
The Three Layers · Layer Two
The Practice: Competency, for the Executive Role
Competency is role-based, and you have a role: synthesis, judgment, communication, and decisions, at volume, under time pressure.
The matrix defines what competency asks in each domain; here is what each asks of you, the leader as practitioner, before the leader as governor. A leader who governs AI without practicing it is governing on hearsay.
Domain 1 · AI Foundations
Match task to tool the way you match assignments to staff
The matrix says: judge which AI approaches fit specific role tasks, and recognize failure modes in real use.
Know which parts of executive work AI genuinely serves, synthesis of long material, stress-testing a decision, first drafts of routine communication, meeting distillation, and which it quietly corrupts. Then know the failure modes as they appear in your tasks: fabricated citations that read perfectly in a briefing; sycophancy, AI flatters the framing you hand it, and executives, already under-contradicted, are the most exposed people on campus to a tool that agrees; false fluency in domains you can't check; and tone-deafness in anything relational.
In practice: you can name, for your own weekly work, three tasks where AI helps, two where it's forbidden, and the failure mode you watch for in each.
Domain 2 · Responsible Use
Your function's requirements are the strictest on campus
The matrix says: apply professional, compliance, and integrity requirements to AI use in your function.
A data floor above everyone else's, personnel actions, legal strategy, labor negotiations, donor intelligence, and board deliberations don't enter tools without agreements covering exactly that class of data, and much shouldn't enter AI at all. An authenticity line, relational writing (condolences, crisis messages, personnel conversations, donor stewardship) must be visibly, genuinely yours, because AI-drafted empathy, discovered, costs more than it ever saved. And a disclosure standard you could defend publicly: if the campus paper reported exactly how your office uses AI, the story would be boring.
In practice: you can state your data floor and your authenticity line in one sentence each, and your office could recite them.
Domain 3 · Applied AI
Redesign the executive workflow, deliberately
The matrix says: redesign real workflows around AI and hold results to a quality bar.
The briefing pipeline: material in, AI produces the first synthesis, a human verifies against sources, you get the distillation plus flagged uncertainties. The decision workflow: before anything consequential, something has argued the other side. The correspondence workflow: triage and first drafts accelerated, your voice and the final call untouched. The quality bar, held without exception: nothing AI-produced leaves the office as fact until verified by a human against a source. A staffer repeating a hallucinated statistic embarrasses themselves; a president citing one to the board creates an institutional incident.
In practice: your office's AI-assisted workflows are documented, your team follows the same verification protocol you do, and you've caught at least one error before it traveled.
Domain 4 · Data Literacy for AI
Two disciplines: inbound and outbound
The matrix says: assess data quality and apply governance requirements in role tasks.
Inbound: interrogate every AI-produced analysis before it enters a decision, where did the data come from, how fresh, what's excluded, and the test that settles it: would we make this decision on this data if a person had brought it? Outbound: apply your institution's data classification to what your own office feeds into tools, you approved the tiers; your office demonstrates them. The executive who asks "what's the source?" in every meeting where an AI analysis appears does more for data discipline than a year of training.
In practice: no AI-generated number has reached your board materials unsourced, and "would we decide on this data?" has become other people's standard question.
Domain 5 · AI Governance & Policy
Where competency and credibility fuse
The matrix says: navigate policy, vet tools, and manage AI risk within the function.
Your office follows the institution's AI policy visibly: the fastest way to kill a policy campus-wide is an executive suite that exempts itself, and everyone will know. Tools your office wants go through the same vetting channel as everyone else's, publicly, because the exception you grant yourself becomes the precedent every dean cites. And you manage your function's own risk profile honestly: the executive office is a shadow-AI risk like any unit, except its data is worse to leak and its example travels further.
In practice: your office has zero exceptions to institutional AI policy, and if asked, your staff would say so without being coached.
Domain 6 · Strategic AI Leadership
Your team is the first institution you transform
The matrix says: champion responsible use and guide a team or unit through change.
Champion responsible use by being watchably competent: disclosing an AI-assisted draft, verifying before relying, declining AI where trust demands the human. Guide your unit through the change: make learning time legitimate, put AI practice on your cabinet's agenda as a working topic, and normalize imperfection, the leader who says "I used AI for this and here's what it got wrong" gives everyone permission to learn out loud. If you can't lead these eight people through it, the campus isn't next.
In practice: someone on your team changed how they use AI because they watched you, and your cabinet discusses its own AI practice as routinely as its budget.
- AI is in your actual weekly workflow, not your rhetoric about workflows.
- You can name your tasks, your forbidden zones, and the failure mode you watch for in each.
- Nothing AI-produced has left your office unverified this quarter.
- Your office would pass the policy audit you didn't announce.
- Someone changed how they use AI because they watched how you do.
The Three Layers · Layer Three
The Fluency Layer
Fluency is the top of the continuum: not using AI better, but setting direction for how your institution adopts and governs it.
The lens over everything below: shared governance. Higher education distributes authority by design. Fluency here includes a capability corporate frameworks never name: knowing which AI decisions belong to shared governance and working through those bodies rather than around them. Direction imposed from the top doesn't just breed resentment, it doesn't stick.
For each of the six domains: what a fluent leader needs to know, what they own, and where the site builds it.
Domain 1 · At fluency
AI Foundations: Assess what's coming
Evaluate emerging capabilities for strategic implication, enough to ask the right questions, never to operate every model. Separate capability from marketing. And own the horizon-scanning discipline itself: name your trusted sources, keep a quarterly rhythm, and require the same of your team.
Know: what these systems fundamentally are, prediction, not understanding, and why capability is jagged: superhuman at some tasks, brittle at neighboring ones. Enough to tell a genuine capability jump from a marketing cycle, and to name the two questions underneath it all, what your degree certifies, and whether graduates leave AI-ready.
You own: the institution's read on what's real. Built in: Module 1, the Quarterly Briefing, Strategic Framing. Agenda: Q1.
Domain 2 · At fluency
Responsible Use: Set the principles, make them hold
You don't just follow responsible-AI principles; you set them. Adopt or adapt an explicit principle set, ratify it through shared governance so it has legitimacy, then do the harder half: ensure it holds at scale, in procurement language, in performance expectations, in what actually gets approved. Principles that never block a purchase or stop a pilot are decoration.
Know: the major responsible-AI principle sets you can adopt or are bound by (your system's own principles, and the UNESCO, OECD, and NIST references they echo), where bias and privacy risk actually originate, and what "responsible use" has to mean concretely, so a principle can block a purchase or stop a pilot, not just hang on a wall.
You own: the institution's values-in-force, and the gap between stated and lived. Built in: Module 5, Trust Leadership. Agenda: Q6, Q9.
Domain 3 · At fluency
Applied AI: Direct investment to where value is real
Pattern recognition across units, not tooling. See where AI genuinely creates value in teaching, research, student success, and operations; direct investment there; decline the rest. Know the build-vs-buy economics: buy the foundation, build the edges, and that total cost runs 3–5x the license once data work, people, and retraining are counted.
Know: where AI genuinely creates value versus where it is a solution seeking a problem, across teaching, research, student success, and operations, and the real economics: total cost runs several times the license once data work, integration, people, and retraining are counted.
You own: the portfolio, including the sunsets. Built in: Modules 2 and 4, Investment Judgment. Agenda: Q8, Q10.
Domain 4 · At fluency
Data Literacy: Govern data as the strategic asset
Your institution's data is not an IT hygiene issue; it is the strategic asset that determines your entire AI ceiling, and someone at your level must own the appetite for data risk. Fund data governance as strategy rather than plumbing, decide which data may be used for which AI purposes (an ethics and mission call), and set risk appetite explicitly, stated before a vendor asks.
Know: your institution's data reality, how much is governed, classified, and actually usable; what FERPA, HIPAA, and existing contracts permit; and why data quality and data rights set the ceiling on everything AI can do for you. The strategy cannot exceed what the data underneath it allows.
You own: the risk appetite, and the decision to fund the unglamorous foundation. Built in: Modules 2 and 3, Agenda Q5.
Domain 5 · At fluency
AI Governance & Policy: Own the whole posture
Decision rights, plus compliance posture (a deliberate stance on FERPA, accessibility, and arriving regulation: where you'll exceed minimums and where minimums suffice) and procurement standards (no training on your data, bias-testing evidence, exit terms, renewal price protection, set once, enforced always, so every purchase doesn't relitigate them).
Know: the regulatory landscape bearing down, FERPA, accessibility/ADA, state law, the extraterritorial reach of the EU AI Act, and the NIST AI Risk Management Framework, what decision rights and procurement standards look like when they actually hold, and where your institution should exceed the minimum versus simply meet it.
You own: policy, posture, procurement standards, and decision rights. Built in: Module 3, Governance Design. Agenda: Q4, Q7.
Domain 6 · At fluency
Strategic AI Leadership: Build the AI-ready institution
The integrating domain. A funded, mission-aligned strategy in plain language (including what you will not do); change led through resistance rather than around it; capability-building for your whole workforce resourced and sequenced; an institution that measures honestly and adapts without lurching. This is where all five other domains compound into direction.
Know: how organizational change and adoption actually happen in a university under shared governance; the workforce and competitive stakes AI raises; and how the other five domains compound, so what you produce is a coherent direction, not a pile of disconnected pilots.
You own: the direction, and the accountability for it. Built in: Modules 1 and 6, Strategic Framing and Change Stewardship. Agenda: Q2, Q3, Q10.
Lead · The Leadership Agenda
The Ten Questions Your Institution Must Answer About AI
Fluency is not knowing about AI. It's being able to answer, for your institution, the questions below, with a defensible answer, a named owner, and a date.
Most cabinets can't answer half of them today. That's not a criticism; it's the starting line. Bring this list to your next leadership meeting and find out which half is yours.
If AI can produce passing work in most of your courses, what are you assessing, and what are you promising employers and graduate schools? This is the existential question, everything else is operations.
Worked in: Ground · DecideYour board will ask this before any other AI question. "Every department is figuring it out" is not an answer; it's an equity problem, because AI readiness will otherwise track existing privilege.
Worked in: Ground · DecideWhich roles change, which skills must be built, and what have you committed to employees? In unionized environments, the answer you don't design will be designed for you.
Worked in: Decide · ActTools, data, procurement, academic policy: for each, can you name the owner, and would everyone in the room name the same person? Ambiguous decision rights are how institutions end up with forty tools and no policy.
Worked in: StructureAI initiatives don't fail on models; they fail on fragmented, ungoverned, poor-quality data. If your institution can't produce one trusted answer to "how many students do we have," it is not ready for the initiative being pitched to you.
Worked in: See · StructureWhat tools are actually in use, what data is actually flowing into them, and which vendor contracts actually address AI? The gap between your policy and your practice is your exposure.
Worked in: SeeA chatbot harms a student in crisis. A deepfake of your president circulates. An algorithm's bias in admissions makes the news. Who speaks, who investigates, who decides whether the system stays on? If the plan doesn't exist before the incident, the incident writes it.
Worked in: See · LeadReal total costs (licenses, data work, people, retraining) against real institutional value, and the harder question: what happens to your enrollment and business model if peer institutions get this right and you don't?
Worked in: DecideAdoption done to people fails; adoption done with people compounds. What have you done to earn the skeptics' confidence, and would the skeptics agree?
Worked in: LeadHonest metrics tied to mission, retention, completion, research capacity, cost, not activity counts. If the answer is "number of pilots launched," the strategy is theater.
Worked in: ActThe questions are universal; the answers are not. "Who decides what" at a 3,000-student community college is a conversation among five people; at a research university it's a negotiation across federated colleges; at a system office it's influence without authority. Where the difference changes the work, the discussion guides say so.
Lead · The Work of AI Leadership
What AI-Fluent Leaders Do
Fluency is the top tier of the framework. For leaders, it shows up as six practices, concrete work you can see a leader doing (or not doing) in any given week. That concreteness is what makes fluency buildable rather than aspirational.
You can explain what AI means for your institution in mission terms, student success, research, equity, financial sustainability, in language your board and faculty senate will act on. Not vendor language. Not technical language. Institutional language.
In practice: you can open a cabinet discussion on AI without deferring to your CIO for the framing.
You can look at an AI proposal and see the exposure the pitch left out: privacy, bias, security, academic integrity, reputation. You know the three questions that reveal whether a vendor has actually thought about higher education.
In practice: you catch a material risk in a proposal before your general counsel does.
You can tell a governance gap from a technology gap. You know who should decide what, tools, data, procurement, academic policy, and how to integrate AI oversight into the governance you already have instead of spawning another committee.
In practice: you can name, for any AI decision on campus, who owns it, and everyone agrees.
You can rank competing AI initiatives by mission fit and readiness, defend the ranking, and the harder skill, end a pilot that isn't working, even a popular one.
In practice: you've said no to a well-pitched initiative and yes to an unglamorous one, and can articulate why.
You can lead a principled, transparent AI conversation with any constituency, including the hostile room. You can hold enthusiasts and skeptics in the same conversation without losing either.
In practice: faculty who oppose AI adoption still say the process was fair.
You can build the coalition, set honest metrics, and sustain momentum past the pilot phase, where most institutional AI efforts quietly die.
In practice: your AI initiatives survive their second budget cycle.
The Operating Principles
The practices are the work. The principles determine whether the work holds up under pressure. Five of them, each visible in how a leader shows up to a meeting.
Every AI conversation starts with an institutional question, not a tool. The wrong question is "what can this AI do?" The right one is "what does our institution need, and is AI the way to get it?"
You cannot govern what you've never touched. Fluent leaders use AI tools firsthand, imperfectly, occasionally embarrassingly, because secondhand understanding produces secondhand judgment.
You treat vendor claims and internal hype with the same scrutiny you'd give a budget projection, while staying genuinely open to what works. Cynicism is as lazy as credulity; both are ways of not thinking.
AI doesn't make institutions effective; people working in well-designed processes do. Buy technology to fix a broken process and you get a faster broken process. Invest in process discipline, tacit knowledge, and psychological safety alongside, and usually before, the tools.
The technology will outrun any policy you write this year. Fluent leaders make decisions built to be revisited, say so publicly, and treat revision as leadership rather than retreat. Permanence is not available; clarity about what you know now is.
Lead · The Path
The Path
Six modules, 4–6 hours total. Every module carries the same Knowledge / Skills / Disposition structure used at every layer of the continuum, plus immediate application. Nothing waits for "someday."
How the six modules cover the six domains
The modules are a sequence, not a domain-by-domain list. There are six modules and six domains, but that is a coincidence of number: the mapping is intentionally many-to-many, not one-to-one. Each module builds capability across two or three domains; read down a column to see which modules develop a given domain. Strategic AI Leadership recurs most, as Fluency's center of gravity.
| Module | 1 AI Found. |
2 Resp. Use |
3 Applied |
4 Data |
5 Gov. |
6 Strategic |
|---|---|---|---|---|---|---|
| M1 · Ground | ● | · | · | · | · | ● |
| M2 · See | · | ● | · | ● | · | · |
| M3 · Structure | · | · | · | ● | ● | · |
| M4 · Decide | · | · | ● | · | · | ● |
| M5 · Lead | · | ● | · | · | ● | ● |
| M6 · Act | · | · | · | · | · | ● |
● = the module builds skill in that domain. Columns are the six domains of the AI Proficiency Continuum; every domain is developed by at least one module.
Knowledge: prediction not understanding; jagged capability; the capability-vs-hype test; why higher education is structurally different; and the two questions underneath everything, what our degree certifies, and whether students leave AI-ready.
Skills (builds Strategic Framing · Domains 1, 6): frame AI for a board in mission terms; apply the capability-vs-hype test in real time; name the two existential questions as leadership-owned.
Disposition: Mission first, technology second.
Practice, In session: draft your three-sentence framing statement; test it against a skeptical trustee. This week: open one leadership meeting with it, and assign an owner to Q1 or Q2.
Knowledge: the opportunity map and the exposure map (seven categories, each with the one question that surfaces it); shadow AI and why crackdowns drive it underground; data readiness as precondition; the compliance floor.
Skills (builds Risk Discernment · Domains 2, 4): assess any proposal for upside and exposure at a glance; ask the three vendor questions that reveal in five minutes whether they've thought about higher ed; surface shadow AI without a witch hunt.
Disposition: Skeptical, not cynical.
Practice, In session: run one live campus proposal through the two-sided assessment. This week: put the three vendor questions to the next pitch, and ask your data leader the "how many students" question.
Knowledge: governance is decision rights, not a committee; the decision map (six decision types × three roles); building on governance you already have; the shared-governance lens; data as strategic asset; procurement standards; the executive sponsor, honestly defined.
Skills (builds Governance Design · Domains 4, 5): complete a decision-rights map and read the disagreements as findings; distinguish a governance gap from a technology gap; state data risk appetite and procurement non-negotiables as leadership positions.
Disposition: Decide provisionally, revise openly.
Practice, In session: fill the decision-rights map individually, then compare; every contested cell is your governance agenda for two quarters. This week: ask your cabinet "who approves a new AI tool here?" and name the executive sponsor.
Knowledge: from pile to portfolio; the five lenses (mission fit, equity, readiness, true total cost at 3–5x the license, risk); build vs. buy; sequencing foundations before features; the discipline of stopping; workforce and competition as strategic constraints.
Skills (builds Investment Judgment · Domains 3, 6): rank initiatives with the five lenses and defend it; estimate true total cost past the license line; conduct a sunset conversation that respects the champion and holds the decision.
Disposition: People make the difference, not the tool.
Practice, In session: rank three to five real initiatives; pick a sunset candidate and draft the rationale for its champion. This week: request the full inventory of AI tools running in your area, the pile, before it becomes a portfolio.
Knowledge: trust as the constraint; setting the principles and making them hold in procurement and approvals; communicating provisionally; followable policy; human oversight specified; incident readiness, drafted before needed.
Skills (builds Trust Leadership · Domains 2, 5, 6): lead a principled AI conversation with any constituency, including the hostile room; apply the human-oversight test to any high-stakes use; own an incident-response posture you could execute tomorrow.
Disposition: Curious enough to use it yourself.
Practice, In session: draft three-minute talking points for your toughest audience; stress-test against the three hardest questions. This week: apply the human-oversight test, which human, what training, what authority, what time, to your highest-stakes AI use.
Knowledge: the honest readiness snapshot; the first 90 days; coalition math; metrics with integrity; the revision cadence; governing, not just observing.
Skills (builds Change Stewardship · Domain 6): produce a one-page roadmap with owners, dates, and honest metrics; score your institution's readiness without flattery; institutionalize the revision ritual.
Disposition: Decide provisionally, revise openly, now institutionalized.
Practice, In session: build the roadmap, three priorities, owners, 90-day milestones, one metric each, and the date of your first quarterly review. This week: put the roadmap on your next cabinet agenda and calendar the first review before the meeting ends.
The deliverables stack: the framing statement (M1) opens the board briefing; the decision map (M3) fills the governance charter; the ranked portfolio (M4) populates the roadmap (M6). And the weekly practices stack too, by the end you've asked the vendor questions, audited your decision rights, seen your real inventory, stress-tested your oversight, and calendared your review.
You haven't taken a course. You've done the first six weeks of your institution's AI leadership work.
Use · Resources
Take These to Your Cabinet
Free, complete, and editable. Every template ships with margin notes for different institutional realities, "if you have no CIO…," "in federated governance…," "at system level…", so a community-college cabinet and an R1 chancellery can both use them as written.
The Module 3 instrument: six decision types, named owners, escalation paths, integration with existing governance.
Institutional AI readiness in the form a board expects: posture, exposure, plan, ask.
The five lenses, mission fit, equity, readiness, true total cost, risk, on one page.
Sponsor, charter, inventory, one visible win.
Success criteria, review date, and the respectful ending.
One per Path module, written so a president can facilitate their own leadership team without anyone's help.
Use · Writing
Writing
Essays, briefings, and working frameworks, from CampusAIExchange.com and the Substack, with the Quarterly Briefing archive below.
Newsletter
The Quarterly AI Briefing for Senior Leaders
What changed this quarter, what it means for your institution, and one question to bring to your cabinet. Five minutes, four times a year. No vendor content, ever.
It's one source for the commitment described in The Foundation, staying current in your scope, with named sources and a rhythm. Build your set; this can be part of it.
Get the BriefingAbout
Built for the Leaders Who Have to Answer for AI
Campus AI Fluency exists because senior leaders are being asked to make consequential AI decisions without a resource designed for their actual role.
Technical training misses the point; generic executive briefings skim the surface. This site defines the work of AI leadership across the proficiency continuum, the knowledge, the skills, the dispositions, and a path for building all three.
About this site
This site is the leadership companion to the Campus AI Proficiency Continuum. That framework maps capability across six domains and three layers, from literacy for everyone, to competency by role, to fluency for those who set institutional direction. This site takes the top layer and makes it usable.
Fluency is the layer the people at the top own and the hardest to find guidance for. Here it becomes concrete: what a leader in each of the six domains needs to know, the questions to be asking, the moves to make in response, and instruments to bring straight to a cabinet or board. The framework tells you what capability looks like; this site helps you build it in yourself and your institution.
If you are new to the framework, start with the system map for the whole picture, then return here for the fluency work.
About the author
Joe Sabado chairs the University of California AI Council's Innovation and Impact Committee, co-leads the EDUCAUSE AI Community Group, and is the author of the Campus AI Framework, the Nine AI Governance Domains, and the Campus AI Literacy model, frameworks in use across higher education.
What distinguishes his counsel is the combination behind it: he runs the enterprise as Deputy CIO at UC Santa Barbara, the systems, data, and integrations of a research university, and brings formal discipline in AI governance (IAPP AIGP), security and privacy (CISSP, CISM), and organizational change (Prosci), grounded in a career that began in student affairs, where his commitment to student success and equity still lives. He evaluates AI the way it actually lands on a campus, not the way it looks in a deck.
About · Speaking & Advisory
Speaking & Advisory
I occasionally accept invitations to speak, facilitate cabinet or board sessions, and advise institutions and systems on AI strategy, governance, and leadership capability.
Recent topics include executive AI fluency, AI governance and decision rights, and building institutional AI capability. If a conversation would be useful for your campus, I'd welcome hearing from you.
Fluency · Domain Profiles
Fluency Domain Profiles
Context · Direction · Role-Specific Knowledge for institutional leaders. These profiles expand the Fluency layer of the AI Proficiency Continuum, the strategic capability expected of presidents, provosts, CIOs, deans, governance officers, and others who set direction.
Each profile provides context on the broader landscape, clear direction for leadership action, and the specific knowledge leaders should internalize. Fluency builds on Literacy (everyone) and Competency (role-based) but shifts focus to institutional stewardship, mission alignment, equity, responsible use, and adaptive strategy. Open any of the six for its full profile.
Designed for executive briefings, governance workshops, capability self-assessment, strategic planning, and professional development.
Fluency · Domain Profile 1 of 6
AI Foundations
Track where capability is heading and its strategic implications.
The domain across the continuum
This page expands the Fluency column, the institutional-leadership expression of the domain, building on the literacy every member of the community needs and the competency each role applies.
Definition
Fluency in AI Foundations means leaders can “read the trajectory” of AI development. In the current landscape of rapid capability evolution (generative models, agentic systems, multimodal approaches, open-source ecosystems), leaders develop contextual understanding to distinguish durable shifts from hype, recognize inherent limits (hallucinations, reasoning gaps, data dependencies, resource costs), and anticipate multi-year implications for higher education’s teaching, research, student success, operations, and public mission.
Significance of the accountable role
Institutional leaders hold primary responsibility for grounding the organization’s AI strategy and decisions in realistic expectations rather than vendor narratives, peer pressure, or short-term trends. This fluency prevents costly misalignments, such as over-investment in soon-obsolete tools or failure to prepare for workforce, pedagogical, or equity disruptions, and enables proactive stewardship of the institution’s long-term viability and values in an AI-influenced world.
Core competencies: what leaders should know
Key responsibilities: direction for leaders
- ·Champion and sponsor regular, evidence-based foresight activities to inform institutional strategy.
- ·Ensure major AI-related decisions, plans, and communications explicitly address realistic capabilities, limits, and future scenarios.
- ·Model critical inquiry and transparent discussion of AI trajectory in leadership forums, board presentations, and community communications.
- ·Direct resources toward sustained capacity for monitoring and interpreting AI developments relevant to higher education.
Success indicators
- ·Institutional AI strategy and planning documents include explicit, periodically refreshed assumptions about AI trajectory and uncertainties.
- ·Leadership decisions demonstrate evidence-based realism (balanced investment rationales that acknowledge both potential and constraints).
- ·Campus stakeholders and governance bodies perceive leadership as well-informed and forward-looking on AI realities rather than reactive or overly optimistic.
- ·Fewer instances of hype-driven initiatives requiring later correction or abandonment.
Critical interfaces
The CIO and central technology/AI leads, academic governance bodies (senate, deans), institutional research and strategic planning offices, research integrity/compliance teams, and external networks (peer institutions, EDUCAUSE, policy bodies).
Common risks if underdeveloped
Decisions based on incomplete or inaccurate understanding of AI’s realistic trajectory, leading to misallocated budgets, unpreparedness for capability plateaus or disruptions, policy gaps, reputational damage from unrealistic public statements, or failure to address equity and workforce implications proactively.
Example scenarios
- ·During strategic planning or budget cycles, leaders evaluate proposals for large-scale AI tools by probing current limits in reliability, integration, data requirements, and long-term viability rather than accepting vendor projections at face value.
- ·In response to faculty or board questions about AI’s future role in teaching or research, leaders provide balanced context on plausible trajectories, associated risks, and institutional preparation steps, guiding informed dialogue and policy development.
Approaches to developing fluency
- ·Participate in or sponsor regular executive horizon-scanning sessions and landscape briefings (internal or with external experts/EDUCAUSE-style forums).
- ·Engage in peer leadership cohorts or site visits with other institutions to discuss real-world trajectory implications and decision trade-offs.
- ·Incorporate AI trajectory questions and scenario planning into existing strategic planning, board retreats, or cabinet discussions.
- ·Commission or review periodic institutional foresight reports and use them to test/update strategy assumptions.
- ·Reflect on past AI-related decisions (successes and missteps) with a focus on what trajectory signals were missed or accurately read.
- ·Build relationships with internal experts (CIO team, faculty researchers) and external networks for ongoing, just-in-time insights.
Fluency · Domain Profile 2 of 6
Responsible Use
Author the institution’s responsible-AI principles at scale.
The domain across the continuum
This page expands the Fluency column, the institutional-leadership expression of the domain, building on the literacy every member of the community needs and the competency each role applies.
Definition
Fluency in Responsible Use means leaders can “set the principles”, define, embed, and enforce institution-wide ethical and mission-aligned commitments that govern how AI is adopted, used, and governed at every scale. This includes operationalizing frameworks such as human-centered design, fairness and equity, transparency, privacy, accountability, safety, academic integrity, sustainability, accessibility, and environmental responsibility.
Significance of the accountable role
Only senior leaders can establish non-negotiable guardrails that survive changes in tools, vendors, or personnel. Without this fluency, responsible-use statements remain aspirational or are inconsistently applied, exposing the institution to equity harms, integrity failures, compliance risks, and loss of trust from students, faculty, staff, and external stakeholders.
Core competencies: what leaders should know
Key responsibilities: direction for leaders
- ·Champion and periodically refresh the institution’s Responsible AI Principles with broad stakeholder input.
- ·Ensure principles are explicitly mapped to procurement, policy, curriculum, performance, and governance processes.
- ·Model principled decision-making publicly, including the willingness to decline tools or uses on equity or integrity grounds.
- ·Establish and resource oversight mechanisms such as AI governance committees, ethics review processes, and impact assessment requirements.
Success indicators
- ·Principles are living documents actively referenced in high-stakes decisions (adoptions, policies, performance reviews, incident responses).
- ·Measurable reduction in documented incidents of irresponsible or inequitable AI use.
- ·Stakeholders can articulate how institutional principles shaped recent decisions.
- ·External reviewers or accreditors note a coherent, operationalized responsible-use posture.
Critical interfaces
General Counsel and compliance; Chief Diversity/Equity Officer; faculty and student governance bodies; HR and labor relations; procurement and vendor management; external civil society AI ethics organizations and peer institution responsible-AI leads.
Common risks if underdeveloped
Performative ethics statements with no enforcement; inequitable outcomes (AI tools that disadvantage marginalized students or staff); academic-integrity or privacy scandals that damage institutional trust and invite regulation or litigation; loss of moral authority when leaders cannot explain principled decisions.
Example scenarios
- ·Evaluating adoption of a predictive-analytics or proctoring tool with known bias risks: leaders require and publicly articulate equity-impact analysis before proceeding or declining.
- ·Responding to a high-profile AI-related academic-integrity incident: leaders apply principles consistently and transparently, focusing on education and systemic improvement rather than solely punitive measures.
Approaches to developing fluency
- ·Sponsor or participate in executive briefings and tabletop exercises focused on responsible-use dilemmas and principle application.
- ·Engage with cross-functional governance bodies or ethics committees to review real or hypothetical cases.
- ·Reflect on past institutional decisions through a principles lens, what was considered and what was missed.
- ·Build relationships with compliance, equity, and academic integrity leaders for ongoing counsel.
- ·Review peer institution policies and case studies to benchmark and adapt approaches.
Fluency · Domain Profile 3 of 6
Applied AI
See where AI creates value; prioritize where to build.
The domain across the continuum
This page expands the Fluency column, the institutional-leadership expression of the domain, building on the literacy every member of the community needs and the competency each role applies.
Definition
Fluency in Applied AI means leaders can “direct the investment”, recognize patterns of AI value and risk across academic and administrative units, prioritize where and how AI should be integrated into core workflows and missions, and allocate resources (budget, people, attention) accordingly. The focus is on strategic portfolio management rather than hands-on tool implementation.
Significance of the accountable role
Leaders control strategic attention and resources. Without this fluency, investment follows the loudest voices, vendor relationships, or lowest-hanging fruit rather than mission-critical value, risk, and equity considerations. This leads to fragmented efforts, duplicated experiments, and missed opportunities for high-impact, well-governed adoption.
Core competencies: what leaders should know
Key responsibilities: direction for leaders
- ·Maintain and regularly refresh an institutional AI investment portfolio view (what is being piloted, scaled, or retired and why).
- ·Establish and enforce clear criteria for moving initiatives from pilot to scaled investment or sunsetting.
- ·Direct capability-building resources toward areas of highest strategic return and managed risk.
- ·Communicate investment rationale, trade-offs, and expected outcomes transparently to internal and external stakeholders.
Success indicators
- ·Clear, documented portfolio of AI initiatives with explicit value hypotheses, risk mitigations, equity considerations, and review cadences.
- ·Resources concentrated on high-impact, well-governed applications rather than scattered low-value experiments.
- ·Regular, defensible sunsetting or pivoting of underperforming initiatives.
- ·Units report that investment decisions feel strategic, transparent, and aligned with institutional priorities.
Critical interfaces
Budget and finance leadership; unit heads (deans, VPs); IT/digital transformation leads; faculty and staff governance; system offices or state higher-ed policy bodies that influence funding.
Common risks if underdeveloped
Fragmented, duplicative, or low-value AI experiments consuming resources without mission return; strategic priorities starved while marginal use cases are funded; inability to demonstrate ROI or impact to boards, legislators, or donors; “pilot purgatory” where initiatives neither scale nor retire.
Example scenarios
- ·Annual budget cycle: every AI-related request must articulate mission alignment, equity implications, integration costs, success metrics, and exit criteria before funding consideration.
- ·Mid-year reallocation: leaders redirect resources from a low-adoption project to a higher-impact initiative (e.g., early-alert or degree-planning) based on usage data and outcome evidence.
Approaches to developing fluency
- ·Review the current institutional AI portfolio with key stakeholders and identify gaps or misalignments.
- ·Participate in or sponsor portfolio review sessions that include value/risk/equity scoring.
- ·Engage with unit leaders to understand real workflow needs versus tool-driven requests.
- ·Study successful and unsuccessful AI investment decisions at peer institutions.
- ·Build internal capacity (e.g., via IT or strategy office) for ongoing portfolio monitoring and reporting.
Fluency · Domain Profile 4 of 6
Data Literacy
Own data strategy and the institution’s appetite for data risk.
The domain across the continuum
This page expands the Fluency column, the institutional-leadership expression of the domain, building on the literacy every member of the community needs and the competency each role applies.
Definition
Fluency in Data Literacy (for AI) means leaders can “govern data as an asset”, treat institutional data as a strategic, high-risk, high-value resource whose quality, classification, access, provenance, and ethical use must be actively stewarded in an AI-augmented environment. AI systems are only as trustworthy as the data they draw upon or are trained with.
Significance of the accountable role
Leaders set the risk appetite, classification schemes, and accountability structures that determine whether data fuels equitable student success and responsible innovation or amplifies bias, privacy harms, and compliance failures. Poor data governance in the AI era creates systemic, often invisible risks that surface in biased outcomes, regulatory violations, or loss of public trust.
Core competencies: what leaders should know
Key responsibilities: direction for leaders
- ·Establish and enforce institutional data classification and handling standards that explicitly address AI use cases.
- ·Require data provenance, bias assessment, and impact review as standard conditions of AI adoption or procurement.
- ·Sponsor data quality and stewardship initiatives that improve AI readiness while protecting privacy and equity.
- ·Model responsible data decisions in leadership communications and personal practice.
Success indicators
- ·Data used in AI systems is classified, documented, and subject to regular review for bias, quality, and appropriateness.
- ·Procurement and project approval processes routinely include data-risk and equity-impact questions with documented answers.
- ·Reduction in data-related incidents or near-misses involving AI tools.
- ·Stakeholders understand and can explain the institution’s data-risk posture for AI.
Critical interfaces
Chief Data Officer / institutional research; privacy and compliance officers; research data management; IT security and identity management; faculty and student data governance committees; system-wide data governance bodies and regulators.
Common risks if underdeveloped
Training or inference on poorly governed or historically biased institutional data; privacy or FERPA violations that surface publicly; inequitable outcomes traceable to data patterns; inability to explain or defend AI-driven decisions to students, families, regulators, or the public.
Example scenarios
- ·Vetting a new predictive-analytics or early-alert platform: leaders require documented data lineage, bias testing on institutional subpopulations, and clear data-retention/deletion policies before approval.
- ·Research data governance for AI projects: leaders set expectations that faculty using institutional or sensitive data in AI work follow classification, consent, and reproducibility standards.
Approaches to developing fluency
- ·Review the institution’s current data classification scheme and AI-related data use cases with data governance leads.
- ·Participate in or sponsor data ethics or bias review exercises for existing or proposed AI initiatives.
- ·Engage with compliance and research data offices to understand pain points and gaps.
- ·Study peer institution data governance models for AI and adapt relevant practices.
- ·Build ongoing relationships with data stewards and privacy experts for just-in-time counsel.
Fluency · Domain Profile 5 of 6
Governance & Policy
Set policy, procurement standards, and escalation paths.
The domain across the continuum
This page expands the Fluency column, the institutional-leadership expression of the domain, building on the literacy every member of the community needs and the competency each role applies.
Definition
Fluency in Governance & Policy means leaders can “own policy & decision rights”, establish, maintain, and enforce the institutional policy architecture, clear decision rights, accountability structures, and compliance posture for AI across the enterprise. This includes creating living policies that evolve with technology and context.
Significance of the accountable role
Policy and governance are the primary mechanisms by which leaders translate principles and strategy into enforceable expectations. Without this fluency, the institution operates with gaps, contradictions, or unenforced rules that create legal, ethical, operational, and reputational risk. Clear ownership prevents diffusion of responsibility and enables coherent response to incidents or opportunities.
Core competencies: what leaders should know
Key responsibilities: direction for leaders
- ·Own the institution’s AI policy portfolio (responsible-use policy, data-classification guide, tool-vetting rubric, syllabus statements, procurement standards, etc.).
- ·Establish clear decision rights and escalation paths for AI-related questions and incidents.
- ·Ensure governance bodies have appropriate authority, expertise, and representation (faculty, staff, students, compliance, IT, academic affairs).
- ·Periodically stress-test policies and governance through tabletop exercises or real incidents and update accordingly.
Success indicators
- ·Comprehensive, coherent, and accessible AI policy suite that is actively referenced and updated on a defined schedule.
- ·Clear, functioning governance bodies with documented decisions, follow-through, and accountability.
- ·Reduced policy gaps or contradictions that create risk or confusion for the community.
- ·Audit, accreditation, or external review findings that affirm a mature AI governance posture.
Critical interfaces
General Counsel; compliance, risk, and audit functions; Academic Senate and faculty governance; student government; IT and information security; system legal/compliance offices, accreditors, and state higher-ed regulators.
Common risks if underdeveloped
Policy vacuums filled by ad-hoc or vendor-driven rules; inconsistent enforcement that appears arbitrary or unfair; compliance failures or accreditation findings; inability to respond coherently and defensibly to incidents or regulatory inquiries.
Example scenarios
- ·Updating academic integrity policy for generative AI: leaders ensure the revision is clear, educative rather than purely punitive, and accompanied by implementation guidance and faculty development support.
- ·Establishing an AI tool review process: leaders charter a cross-functional body with defined decision rights and publish transparent criteria and timelines for the community.
Approaches to developing fluency
- ·Review the current AI-related policy inventory with General Counsel and compliance leads to identify gaps.
- ·Participate in or sponsor tabletop exercises or scenario planning focused on policy application and gaps.
- ·Engage governance bodies (senate, committees) in collaborative policy development or review.
- ·Study peer and system-level AI policies and governance models for adaptation.
- ·Build relationships with legal, compliance, and academic governance leaders for ongoing partnership.
Fluency · Domain Profile 6 of 6
Strategic Leadership
Set strategy, lead change, and build an AI-ready organization.
The domain across the continuum
This page expands the Fluency column, the institutional-leadership expression of the domain, building on the literacy every member of the community needs and the competency each role applies.
Definition
Fluency in Strategic Leadership means leaders can “set direction”, articulate a coherent, mission-aligned AI vision; make deliberate adoption or refusal decisions; resource capability-building; manage systemic risk; and lead the cultural and organizational change required for responsible AI integration at scale. This is the capstone domain that integrates all others.
Significance of the accountable role
Only senior leaders can align AI strategy with institutional mission, values, and long-term sustainability; balance competing stakeholder interests; and model the adaptive, evidence-based leadership the entire community needs. Without this fluency, AI efforts remain tactical and fragmented, or the institution drifts reactively in response to external pressures.
Core competencies: what leaders should know
Key responsibilities: direction for leaders
- ·Articulate and regularly refresh the institution’s AI vision and strategy, including explicit boundaries of what the institution will and will not do.
- ·Make and defend high-stakes adoption, investment, and policy decisions with evidence and transparent rationale.
- ·Sponsor and resource institution-wide capability-building sequenced by risk and strategic value.
- ·Lead cultural change by modeling responsible use, adaptive learning, transparent communication, and human-centered leadership.
- ·Establish and oversee the AI operating model and measurement framework that enables learning and adaptation.
Success indicators
- ·A living, funded AI strategy exists and is widely understood, referenced, and aligned with the institutional strategic plan.
- ·High-stakes decisions are evidence-based, documented, and revisited on a defined schedule.
- ·Capability-building reaches the right roles at the right depth and is measured for behavior and outcome change.
- ·The institution demonstrates adaptive capacity, learning from pilots, incidents, and external shifts without disruption to core mission.
- ·Stakeholders report clarity of direction, psychological safety to raise concerns, and confidence in leadership’s handling of AI.
Critical interfaces
President’s cabinet and executive team; Board of trustees/regents; faculty and staff senates; student leadership; system leadership, state/federal policy bodies, peer networks, accreditors, donors, and foundations.
Common risks if underdeveloped
Strategic drift or purely reactive posture (“AI strategy by inbox”); loss of trust from inconsistent or opaque decisions; capability investments that fail to produce institutional outcomes; reputational or mission damage from poorly governed or inequitable AI adoption at scale.
Example scenarios
- ·Board retreat or strategic planning cycle: leaders present a clear AI vision with prioritized domains, explicit risk appetite, resource requirements, and multi-year milestones.
- ·Major incident (e.g., biased AI tool affecting admissions, advising, or research): leaders respond with transparent review, principle-aligned corrective action, public learning, and systemic improvements rather than defensiveness.
Approaches to developing fluency
- ·Lead or actively participate in strategic planning processes that explicitly address AI.
- ·Engage in executive peer cohorts or national higher-ed AI leadership forums for shared learning and benchmarking.
- ·Reflect regularly on institutional AI decisions and outcomes with a strategic lens (what worked, what didn’t, why).
- ·Build trusted relationships across academic affairs, IT, student affairs, compliance, and faculty governance.
- ·Commission or review institutional AI maturity or readiness assessments and use results to guide direction.
Use · Glossary
The terms you'll hear, defined for the desk
Not a technical dictionary. These are the words that surface in a vendor pitch, a board question, or a governance debate, defined at the depth a leader needs to govern credibly, and no deeper. The test is never whether you can build it; it's whether you can question it.
A companion to the plain-language "what they'll say / what it means" translations on The Foundation.
Definitions are pitched for governance, not engineering, enough to ask the right question and read the room, never to operate the system yourself.