Buying AI Tools Isn’t Building AI Capacity

By Claire L. Brady, EdD

This post is part of a series exploring Achieving the Dream’s Creating the AI-Enabled Community College framework through a leadership lens. While ATD is focused on community colleges, the report’s core questions about strategy, governance, culture, workforce readiness, professional learning, and student success apply across institution types. Throughout the series, I’ll examine what its eight action areas look like in practice and what it takes to move from scattered AI activity toward intentional, sustainable institutional change.

Purchasing an AI platform can happen relatively quickly. A contract is negotiated, licenses are distributed, and a campus announcement lets everyone know that a new tool is available.

Then leaders wait for transformation to happen.

Some employees begin experimenting immediately. Others log in once and never return. Many are interested but unsure how the tool applies to their work. Faculty wonder how it should change teaching and assessment. Staff are told the technology will make them more efficient, but no one has helped them reconsider the processes surrounding it. The institution has invested in access without building the conditions that make access useful.

This is why buying AI is not the same as building AI capacity.

Capacity Requires More Than Training

Achieving the Dream separates staff capability and faculty professional learning into two action areas, but the underlying message is consistent: institutions cannot simply give people another technology and expect them to figure out how to use it well.

For staff, ATD calls for more than technical instruction. It emphasizes identifying where AI can strengthen student success, redesigning systems and processes, providing significant support, and bringing cross-functional teams together so new approaches can be integrated into real workflows without losing necessary human involvement.

Faculty need the same depth of investment. A demonstration of features may show someone what a tool can do, but it rarely answers the questions that matter most: How does this affect learning? When does AI support students, and when does it short-circuit the work they need to do? What should change in an assignment? What remains distinctly human about teaching?

A one-time workshop may increase awareness. It does not automatically create confidence or sound judgment.

From Literacy to Institutional Practice

In AI with Intention, I describe AI literacy as capability that develops through repeated engagement, application in context, and reflection—not information delivered in a single session. Effective learning is role-specific, practice-based, ongoing, and owned across the institution rather than treated as generic technology training provided only by IT.

I think of the progression this way:

AI literacy → AI confidence → AI capability → Institutional practice

Literacy gives people a foundational understanding of the opportunities, risks, and limitations of AI. Confidence develops when they have safe opportunities to experiment, ask basic questions, and acknowledge what they do not yet understand. Capability comes through using AI in the actual contexts of their work and learning when to trust it, challenge it, redesign around it, or decline to use it. Institutional practice emerges when those capabilities are supported by expectations, workflows, governance, time, and resources.

Institutions often try to jump from the first stage to the last. They offer an introductory workshop and then expect employees to redesign services, courses, and processes on their own. That is not capacity building. It is unfunded implementation work.

Invest in the Conditions for Learning

The institutions making meaningful progress are creating more sustained structures: discipline-based faculty cohorts, role-specific learning pathways, communities of practice, coaching, office hours, internal knowledge bases, and protected time to experiment. In one example from AI with Intention, an institution replaced generic training with monthly learning cohorts, role-based pathways for staff, an AI Fellows program, and time built into workload expectations. The result was not simply increased use; faculty began redesigning assignments thoughtfully, and advisors better understood how to use AI without making student interactions less personal.

That last point matters. The goal is not to produce more AI users. It is to help people improve their work while protecting the judgment, relationships, and expertise that make the work valuable.

Leaders also need to recognize that learning AI creates workload before it saves any. ATD specifically recommends examining both the efficiencies AI may generate and the additional demands created by training, implementation, and professional development. Asking already-stretched employees to build new skills without time, support, or relief elsewhere is not an investment in readiness. It is another expectation layered onto the same people.

Institutions cannot purchase their way to AI readiness. Tools may make change possible, but people turn possibility into practice.

A Question for Leaders

• When you compare what your institution is spending on AI platforms with what it is investing in people’s time, learning, support, and process redesign, what does that balance reveal?

A license gives someone access to AI. Capacity gives them the ability and confidence to use it wisely.

This image was created using ChatGPT