The AI Literacy & Professional Learning Framework.

The AI Proficiency Continuum, operationalized through professional learning curriculum & offerings.

AI proficiency is the demonstrated ability to use AI in practice across a continuum — from foundational literacy (understanding and critical evaluation), through role-specific competency (responsible application), to advanced fluency (creative, adaptive integration and collaboration with AI systems).

The framework turns that continuum into action — role-based learning pathways, curriculum, and offerings that build capability layer by layer. It rests on one principle: AI should augment human capability and judgment, not replace them.

Start with principles. The principles are the lens every capability and every offering is designed through — the ethical and mission-aligned commitments that decide what “responsible” use looks like at each layer. See how they drive the continuum →

Short on time? Read the Executive summary — the whole case in three minutes.

From framework to practice, in four moves.

The framework defines what capability to build. The operating model is how you stand it up and keep it running — a repeatable loop any institution or unit can pick up: Assess → Plan → Deliver → Measure, then around again.

Stand up a program.

This is the provider side — for the people who sponsor, lead, design, and run the programs, not the people receiving them. The four moves are the same; you're just choosing the scale.

Pick a scope — the team and the playbook below adjust to fit.

{{ pbBlurb }}

The operating model, as a running order for {{ pbScopeLabel }}.

Program briefs.

Sponsor-ready one-pagers for specific program patterns. Each says what it is, why it's needed, how it runs, what it covers, what it costs, and exactly what you'd ask an executive sponsor to approve — a summary you can take into the room.

These are representative patterns, not a fixed catalog — starting points to adapt, combine, or rename for your institution.

The capability map.

Two things define this work: what capability you're building (the three layers) and how you build it (the four moves of the operating model). This is both at once — read a row to follow one layer through the loop, or a column to see one phase across all three layers.

Click a layer to read it in full. The loop repeats — each Measure feeds the next Assess.

How to actually stand this up.

A framework isn't a plan. This is the arc a seasoned team runs — six phases from first conversation to a self-sustaining program. Each phase names the questions to ask, who to consult, and the artifacts you leave with.

The four-move operating loop runs inside each phase; this is the sequence that carries an institution through the first full turn.

Where does your institution stand?

A ten-part readiness check across the foundations and each audience. It scores your maturity, maps your capability gaps, and points you to the right first phase and programs.

Answer as your institution is today, not as you hope. Best done with a few stakeholders in the room. Saves on this device.

Whose voice is in this?

Capability-building succeeds or fails on stakeholder alignment. This maps every group, the interest they bring, and their role across the six phases — so the plan reflects the whole institution, not one office.

Read it as a working RACI: who's Accountable, Responsible, Consulted, or Informed at each phase.

The artifacts you leave behind.

Every phase of the arc produces a concrete deliverable. These are the working documents — each meant to be co-produced with stakeholders, not handed down. Use the structure below as your starting outline.

Five of these open as working, fillable artifacts — fill them in here and your entries are saved on this device. The business case and measurement framework open their full interactive builders.

What it costs to build capability.

The first question leadership asks is "what does this cost?" — and the honest answer is it depends on the shape you choose. This page gives planning ranges, not quotes. Treat them as orders of magnitude to pressure-test a budget, then localize.

Ranges reflect typical U.S. higher-ed programs and assume you build on existing staff and platforms where possible. Your context — institution size, wage bands, existing licenses — moves these substantially.

Enterprise AI licensing is typically separate: plan on a per-user annual cost for licensed seats, negotiated in volume. Many institutions start with a limited licensed pool rather than universal seats.

The obligations that don't bend.

Capability-building is optional; compliance is not. This page translates the legal landscape into what it means for AI use on campus. It is orientation, not legal advice — confirm specifics with your counsel and compliance offices.

See also the regulatory landscape for the broader policy picture, and the policy & language pack for ready-to-adapt language.

Legal minimums protect the institution; equity protects people. Both belong in every capability program.

Language you can start from.

The most-reused assets in this field are the ones nobody wants to write from scratch: a responsible-use policy, syllabus statements, a data-classification guide, and a tool-vetting rubric. Copy any block, adapt the bracketed placeholders, and route through your own governance and legal review.

Starting templates, not finished policy. Bracketed [PLACEHOLDERS] must be localized. Pair with compliance & accessibility.

{{ pc.text }}

{{ artPurpose }}

A starting structure, not a form to satisfy. Fill it in as you run the engagement — everything you type is saved on this device.

Too large, too distributed, and too varied for one-size-fits-all.

Higher education spans research universities, community colleges, health systems, and national labs — each with its own culture, roles, and AI direction. No single training fits all of them — which is exactly why the model is layered by role.

A vast, varied, and shifting institutional landscape.

The 3.66M-person workforce sits inside roughly 3,900 degree-granting institutions — no two alike in mission, size, or resources. That diversity is exactly why a shared framework must be adopted locally.

The AI workforce question higher ed isn't asking.

The familiar question is "how do we prepare students for the AI-era workforce?"

Necessary — but incomplete. It misses the other half: what are the AI workforce implications for higher education as an employer? The people who run advising, finance, research administration, IT, and operations are already using AI, often without training or guardrails.

Staff are already using AI — without institutional guidance.

This isn't a future-readiness exercise. AI is in the work today; the question is whether it's used safely and well.

Eight segments. Three in the very-high band.

AI exposure is highest exactly where higher ed's operational workforce sits — administration, finance, and technical roles.

The retention floor is already fragile.

Two futures for the higher ed workforce.

Staff are already stretched — 51% work beyond full-time hours, 61% have taken on duties outside their job description, 53% have absorbed work from departed colleagues. Whether AI relieves that strain or worsens it depends on having a competency strategy.

AI capability is becoming a compliance obligation.

Regulation is shifting from voluntary principles to enforceable rules — and some now mandate AI literacy directly. Requirements differ by jurisdiction and sector, so institutions operating across state or national borders face overlapping obligations. The framework offers one structured way to respond.

Informational only — not legal advice. Obligations vary and evolve; consult institutional counsel and compliance teams for specific requirements.

A crowded field, converging on a few ideas.

Since 2016, dozens of AI literacy, competency, and fluency frameworks have appeared — from conceptual research models to national and institutional standards. They differ in audience and depth, but a shared spine has emerged. This is a brief synthesis and a component-level comparison of the most influential.

Which components each framework covers.

Eleven influential frameworks mapped against eight recurring components. ● covered · ◐ partial · – not addressed.

What the field agrees on — and where it's thin.

This framework adopts the field's convergent spine — foundations, use, evaluation, and ethics — and extends it in the three areas the landscape leaves thin:

The field is rich, and the frameworks above are genuinely valuable — this work builds on them, not against them. The intent is not to add one more competing model, but to connect what already exists and extend it to where higher education actually needs it.

Enablement is a shared commitment — not a program handed down.

Institutions provide the strategy, platforms, and programs. But building capability is also each person's responsibility. Real progress happens when the individual, their manager, the unit, and the institution each hold up their end — responsibilities that overlap and reinforce one another.

Value that compounds — for the person and the institution.

AI enablement pays off on two ledgers at once. Individuals gain capability, confidence, and career resilience; institutions gain productivity, lower risk, and a workforce ready for what's next. The two reinforce each other.

Match the format to the outcome.

No single format builds capability on its own. Each offering type suits a different depth of learning, audience, and goal — and each should be measured against outcomes, not just attendance. Below: what each is for, how to know it's working, and what to watch.

An offering type is one piece of a five-part pipeline. The continuum becomes real when an intended outcome drives a curriculum, delivered through an offering type, experienced as realistic learning, and held by ongoing support and recognition.

Participation is voluntary and invitational. A sensible first step depends on where you sit:

Designed to support faculty, staff, and students in navigating AI — not to impose compliance-only training.

Higher education is not a typical enterprise.

Capability programs borrowed from corporations rarely transplant cleanly. Universities run on shared governance, distributed authority, and academic freedom — features, not bugs. A framework for higher ed has to work with that grain, not against it.

One tension deserves to be named plainly, because it stalls real programs: ownership is genuinely contested. AI literacy sits across lanes that shared governance keeps deliberately separate — and no single office can simply claim it. Student AI literacy is the sharpest case.

We’re not really enabling AI. We’re enabling our communities.

It is tempting to frame this as an AI project. It isn’t. AI is the occasion; people are the point. The goal is a community of students, faculty, researchers, and staff who can navigate a changing world with judgment and confidence — AI is simply one capability among the many that serve that mission.

What it looks like in practice.

Six illustrative scenarios — each a real kind of problem higher education faces, and how the framework addresses it. They show the value in concrete terms: the need, the response, and what changed.

Eight shapes a faculty cohort can take.

When you design a cohort offering, you're choosing a shape — a focus that trades depth, reach, cost, and durability against one another. These eight archetypes name the common ones.

A snapshot of programs in the field.

A non-exhaustive snapshot of AI programs and initiatives that publicly exist across higher education — spanning access at scale, curriculum-wide literacy, faculty development, and shared resources. They show the range of activity underway; they are not a ranking, and they don’t necessarily align with this framework.

How do you know it's actually working?

The hardest question in capability-building is also the most important: what changed because of this? Sign-ups and completions are easy to count but say little. Real impact means climbing from activity to capability to outcomes that matter for the mission.

Adapted from established training-evaluation models. Higher rungs are harder to measure but far more meaningful — most programs stop at rung 1–2; aim to reach 4–5.

"Working" looks different for each format. Select an offering to see the signals that matter most for it.

Who it takes to build and run the program.

Programs like these don't run on a single hire. They need sponsorship from the top, dedicated program leadership, people who design and deliver the learning, and specialists who keep it safe and measurable. These are functions, not necessarily headcount — at a small institution one person may cover several; at a large one each may be a team.

From framework to action: a starter playbook.

What it actually takes to stand up enablement — the investments, people, tools, and technology to plan for, plus a phased checklist to start. Treat it as a menu scaled to ambition and budget, not a mandatory bill of materials.

The real budget lines behind a program — a mix of people, licenses, content, and time.

Illustrative categories, not endorsements — choose per institutional context, security review, and existing contracts.

The minimum functions to cover — often one person wearing several hats at first. See Roles & team for the full map.

A pragmatic sequence — mapped to the roadmap's Define → Develop → Implement phases.

AI literacy is one pillar of a bigger picture.

Capability-building is necessary, but not sufficient on its own. Building capability only creates institutional value when it sits alongside mission, principles, policy, governance, and operations. The Campus AI Framework maps that whole: nine institutional pillars for responsible AI, of which this site's work — Pillar 5 · Engagement & Collaboration — is one.

Seeing literacy in this context matters: without policy and governance, capability creates risk; without readiness and community, programs stall; without mission, AI drifts. The pillars work together.

Applied across four domains: teaching & student success · research & innovation · operations & infrastructure · student & community engagement.

The values every layer is built on.

The framework repeatedly points to "responsible use." These are the principles behind that phrase — a shared set of commitments that shape how AI is taught, adopted, and governed at every layer. They are widely held across UNESCO, the OECD, and the NIST AI RMF, and institutions should adapt them to their own mission and values.

The principles are the shared values; the six domains are where they get exercised. Most principles touch several domains — the map below shows where each is chiefly lived out.

Responsible use isn't only whether to use AI — it's how a person works alongside it. A useful shorthand from the field frames four habits of good human-AI collaboration:

These principles aren't novel — and that's the point. They converge with the major responsible-AI frameworks in use across higher education and government, so adopting this framework doesn't mean choosing against the standards your institution already references. The crosswalk below shows where each set's principles land.

A convergence map, not an endorsement or a claim of equivalence — always adopt the specific principle set your institution and system are bound to. See sources & references.

The framework in brief.

A three-minute synthesis for decision-makers. Everything below is expanded elsewhere on the site.

Tip: use your browser’s Print (⌘/Ctrl-P) to save this page as a one-page PDF.

Start here — by who you are.

This site is deep. These are short paths through it, depending on why you’re here. Follow one — you don’t need to read everything.

About this website.

This is a working reference for building AI capability across a higher-education institution. It defines a shared model of what AI proficiency looks like — a continuum from Literacy to Competency to Fluency across six shared domains — and then lays out how to turn that model into professional learning and put it to work at the system, campus, and program levels. One idea organizes everything: capability should be defined once, shared as a common language, and built through role-based learning that rests on responsible-AI principles.

Every top-level section, generated from the site's own navigation. Select any to open it.

Key terms, defined.

The framework depends on precise language. These are the core terms as used across this site.

Now what?

This framework is a living, evolving model, strengthened by the higher-education community. There are a few ways to move it forward.

There’s a version of this for every institution.

Approaches span a wide range — from nascent to mature, and from a single volunteer to a well-funded enterprise team. The framework is the same at every point; only the footprint changes. Find the profile closest to yours and start there.

Beyond size and maturity, programs differ by where they live — a single unit going it alone, or a system and its campuses sharing the work.

Why a framework — and why this shape?

AI capability is being built everywhere at once, unevenly and in isolation. Without a shared structure, effort fragments: the same trainings get rebuilt, gaps go unnoticed, and no one can see the whole. A framework is not bureaucracy — it is the common language that lets distributed effort add up to something coherent.

The three layers exist because AI need is not uniform. A universal baseline everyone shares; deeper, applied skill where a role demands it; and strategic capacity for the few who set direction. One tier would either overwhelm most people or underserve those who need depth.

Know where you stand — as a program and as a person.

Assessment works on two levels. A maturity model gauges how developed the enablement program is; a skill assessment gauges individual capability against the framework. Each offering also carries its own way of checking that learning happened.

A staged model for how far an institution's AI-enablement effort has developed. Most start at Ad hoc; the goal is to move deliberately toward Embedded.

Capability is assessed against the three layers — through practical demonstration, not seat time. Self-assessment sets a baseline; evidence confirms it.

Map offerings to what they build and how each checks learning — so the program covers all layers with no gaps or redundancy.

{{ audience.headline }}

{{ audience.intro }}

Addressing the resistance.

Any framework meets skepticism — some of it well-founded. Naming the common objections directly, and answering them honestly, is part of earning adoption. Here are the ones decision-makers hear most.

One foundation for all; depth where roles require it.

The framework has two axes. The first is depth — three cumulative layers of capability. The second is substance — six domains that run through every layer. Cross them and you get the capability matrix — the framework in one view.

Start with the depth axis: the three layers are cumulative, not separate tracks. Everyone builds Literacy. Competency is added where a role calls for it. Fluency is for institutional leaders who set AI direction.

The six domains are the shared foundations — they apply at every layer, with emphasis shifting as depth increases. They answer what capability is about, where the layers answer how deep it goes.

AI Literacy

Know enough about AI to use it safely and question it well.

The shared foundation for the entire campus community. Covers what AI is, its limits and ethics — plus the judgment to question outputs and weigh risks like bias, privacy, and accuracy. Its larger aim is critical thinking about AI that outlasts any specific tool.

Grouped under the six competency domains — the same foundation runs through every layer; at Literacy the emphasis is awareness and safe basics.

What literacy asks of every person. Knowledge is what they understand; skills are what they can do; disposition is the stance that keeps AI use safe and thoughtful.

The six competency domains are the shared foundation across all three layers. At the literacy layer, each is expressed as awareness and safe basics — AI Foundations and Responsible Use are its center of gravity.

AI Competency

Use AI well in the actual work of your role.

Defined by role and operationalized through the specific tasks and workflows of each domain. It includes the ability to use AI tools effectively and responsibly, evaluate outputs critically, integrate AI into existing workflows, and apply responsible-use principles in daily practice.

The same six domains, now at role depth — applied to the tasks, workflows, and standards of the function. Pick a role to see how each domain lands for it.

What competency asks within a role. Knowledge is what a practitioner understands; skills are what they can do in the work; disposition is the stance that keeps AI-assisted work accountable.

The six competency domains are the shared foundation across all three layers. At the competency layer, each is expressed as applied, role-specific practice — Applied AI and Data Literacy are its center of gravity.

AI Fluency

Set direction for how the institution adopts and governs AI.

Adaptive expertise — the ability to evaluate emerging AI capabilities, assess institutional fit, make informed adoption decisions, sponsor initiatives, and shape organizational AI strategy. This is a capacity to build toward, not one to assume is already in place.

The same six domains at institutional depth — where governance, strategy, and stewardship become the emphasis.

Fluency is measured less by what an individual can do than by what the institution can do because its leaders are fluent. A fluent leadership cohort produces:

The knowledge, skills, and disposition that underwrite those outcomes. Knowledge is what a fluent leader understands; skills are what they can do; disposition is the stance that holds it together under uncertainty.

The six competency domains are the shared foundation across all three layers. At the fluency layer, each is expressed as direction-setting rather than hands-on practice — Strategic AI Leadership is its center of gravity.

Six domains of AI capability.

These six domains are the shared foundation of the whole framework — the one axis that holds constant while the three layers (Literacy → Competency → Fluency) change how deep each goes. Everyone, in every role, works across all six; what shifts is the emphasis.

This page defines each domain. To see how one plays out at each depth — the full crosswalk — open it in the capability matrix.

From principle to curriculum — and role.

The spine of the continuum is the intended outcome — the demonstrable capability a person should reach. Everything else is built around it: domains say where the capability lives, curriculum is derived to reach it, and offerings deliver it — with the outcome shifting by role and layer.

Responsible-AI principles are one useful lens onto this: each value points to the outcomes that would honor it. This page follows that lens — starting from a principle and tracing it to the outcomes, curriculum, and offerings that make it real.

All roles — the institution-wide view: how each principle deepens across Literacy, Competency, and Fluency. Pick a role to see the intended outcome and the curriculum that reaches it.Showing {{ cwRoleLabel }} — the intended outcome for this role and the curriculum designed to reach it. The fuller six-domain profile is on the framework by role.

Principles shown are the UC Responsible AI Principles (UC AI Council / UCOP ECAS). The domain, curriculum, and layer mappings are illustrative — a way to make the principles teachable, not an official UC position. See sources & references.

The framework, for your role.

The six domains are shared by everyone — but what they ask of you depends on your role. Pick a role to see, under each domain, the knowledge it needs, the skills it builds, the disposition it calls for, and the curriculum topics that could get someone there.

Literacy stays universal for everyone; role differences show up mostly at Competency and, for leaders, Fluency.

Every domain applies to this role — what shifts is how each shows up in the work.

A competency here is expressed as knowledge, skills, and disposition, plus example curriculum topics, at the depth a given role needs. Illustrative, not prescriptive — adapt to your institution's roles and programs.

This matrix adapts an established pattern — proficiency levels crossed with competency domains, as seen in Bloom's taxonomy, the EU DigComp framework, and UNESCO's AI competency work — to higher education, extending the top of the ladder to institutional leadership and governance. It is a synthesis of the field, not a replacement for it. See sources & references.

Where do you fit in the framework?

Everyone starts with Layer 1. Select the description closest to your role to see every layer that applies to it — and where your role's focus sits.

How capabilities are developed over time.

The professional-learning dimension describes how each layer can be developed through existing institutional channels and targeted new offerings. It is guidance — decisions about specific programs and formats remain with local leadership.

Two different — but related — topics.

The framework and its implementation are distinct. The framework is a shared guiding structure — a common language and set of capabilities. Implementation is the local work of adopting it. Keeping them separate lets the framework stay stable and sector-wide while adoption flexes to each context.

Implementation can range from informal to formal — scaled to the institution's context, capabilities, and need. The framework is the same; the footprint flexes.

The framework is the same everywhere; the realistic footprint differs sharply by institution type. A few illustrative patterns — adapt to your own reality:

These are tendencies, not rules — a small college may run a formal program, and a large system may start with one embedded session. Use them to locate your realistic starting footprint, not to box yourself in.

What to run centrally, what to leave to campuses.

In a multi-campus system — a central or system office plus distinct campuses — the framework is one shared language, but delivery is split. The system standardizes what benefits from consistency and scale; each campus owns what depends on local context and relationships.

The test for each program: does it get better by being the same everywhere, or by fitting the place? That answer sorts it into one of three modes.

Generic multi-campus model — adapt the central/campus split to your own system's governance, funding model, and shared-services arrangements.

Define, develop, implement.

A realistic sequence any institution can follow to move from definition to systemwide practice.

Strategic oversight, operational ownership.

The framework is owned and resourced centrally for consistency and sustainability, with a clear split between strategic oversight and day-to-day operation.

Adopting the framework: three actions for leadership.

A small, time-bound workgroup to advance the framework.

Six steps for higher education.

A practical sequence for turning the framework into workforce capability.

Sources & references.

Figures on this site draw on national datasets, sector research, and recognized AI governance frameworks. Readers should consult the cited sources directly.