“We Don’t Have the Capacity for AI”: A Reality for Many Community Colleges

TL;DR

Community colleges have lost more staff than almost any other institution type and recovered less of it, which means the honest objection to any new initiative, including AI, is usually “we don’t have the bandwidth.” That same understaffing is exactly what makes community college retention strategies harder to execute, since a team without time to review at-risk flags can’t act on them. This piece looks at why AI agents built for that gap don’t ask a stretched team for more time. They give it back.

• Community colleges lost ~13% of staff, 2020–2022
• Understaffing breaks retention strategies
• AI agents give time back, not more work
• Same team, no new headcount

Natalie Hayes runs student success at a two-year college in the Midwest. She also, unofficially, runs half of academic advising, because two advisor positions have sat unfilled since spring.

Her Tuesday looks like this: a budget meeting at 9, a backlog of 40 unanswered student texts by 11, a call with a vendor pitching “an AI solution” at 2, and a stack of early warning flags she hasn’t had time to review since last week.

When the vendor asks if she’s ready to add AI to her strategy for next year, her honest answer is no. Not because she doesn’t believe it works. Because she doesn’t have the hours to learn one more system.

Why Retention Suffers When Teams Are Already Stretched

That answer is common, and it’s not a personal failing.

Community colleges lost a disproportionate share of their workforce during the pandemic and have recovered less of it than four-year institutions. A Chronicle of Higher Education analysis found that these institutions lost roughly 13 percent of their staff nationally between January 2020 and April 2022, and unlike four-year institutions, which mostly recouped those losses, two-year colleges have lagged in their recovery.

Burnout was one of the drivers researchers pointed to, alongside better pay elsewhere and the pull of remote work, and it’s still showing up the same way today: staff picking up a second job’s worth of work with no second paycheck.

That stack of flags on Natalie’s desk isn’t a side effect of being busy. It’s the actual mechanism by which understaffing turns into a student who stops out. Early warning systems exist precisely because a warning sign caught in week three is a conversation, and the same sign missed until week nine is a student who’s already gone.

Most retention strategies, at a community college or anywhere else, depend on someone having the time for early intervention before a student disengages for good. This is what the failure funnel looks like from the inside: not one big failure, but a hundred small ones nobody had time to catch. Student retention and student engagement aren’t separate problems here.

Montcalm Community College, a rural two-year college in Michigan, is one example of a team that closed that exact gap without adding headcount. When someone at a two-year institution says they don’t have the capacity for a new initiative, they’re usually right, given how most new initiatives work.

Here’s where the objection needs a second look. Most new tools ask a stretched team to learn a new system, manage a new inbox, and add a new task to an already full day. That’s a fair thing to say no to. The math for a new initiative rarely changes the equation. It just adds another variable to it.

Capacity = Staff Time × Efficiency

Staff time is the side of that equation nobody can grow right now. Hiring freezes, unfilled positions, and budgets that don’t stretch mean the number of hours available hasn’t gone up in years for most community colleges, even as enrollment and student need have.

Efficiency is the only lever left, and it’s the one AI agents are built to move. The question worth asking about any new tool isn’t “do we have time to add this.” It’s “does this tool take work off the plate, or does it add to it.”

Most tools add to it. That’s why the objection exists in the first place, and it’s a reasonable one to hold onto until a tool proves otherwise.

What Protects Community College Student Success

An AI agent built for a team like Natalie’s doesn’t ask her to learn a new inbox. The Inbox Assistant, for example, works inside the inbox her team is already using, answering routine questions, tagging and routing messages, and covering the evenings and weekends when nobody’s watching the queue.

Her team doesn’t add a task. They get back the hours currently spent triaging the same handful of question types over and over, so the time they do have goes toward the early alert flags sitting in that pile, which is where community college student success actually gets protected or lost.

It’s the same hours that determine retention rates and, further down the line, completion rates, since a student who’s caught early is a student who’s more likely to still be enrolled next semester.

This is where “we don’t have the capacity for AI” and “AI can build capacity back” stop being in conflict. They’re describing the same institution from two different points in time.

The first sentence describes a team that’s out of hours to give to something new. The second describes what happens when the new thing hands hours back instead of asking for them. Same team. No new headcount. More room for the work that actually needs a person.

None of this erases what community colleges are up against. Working adults, first-generation students, and returning learners often need more support, not less, and a smaller staff can’t stretch to meet that on its own. But the answer was never going to be “hire more people,” because that lever hasn’t been available for a while now.

The answer has to be making the hours a team does have go further, and that’s a different kind of solution than the one most vendors are pitching.

Your team didn’t fail to keep up. The math never worked in the first place, and it’s not going to fix itself by asking people to do more with the same 24 hours in a day. The tools that are actually worth a stretched team’s time are the ones that shrink the pile instead of adding to it.

Read next: For the fuller version of this argument across every institution type, see The Higher Ed Capacity Crunch: Why Technology Decisions Are Now Workforce Decisions.

---

Author: