A student uses generative AI in an English class on an assignment that asks them to draft a routine professional email. The assignment allows it, the information is not sensitive, and the student reviews the final message before submitting it. Later that day, the same student needs to write an email at work. AI could help with this message, too. But now the email includes information about a patient, customer, employee or internal workplace matter.
The writing task has not changed very much. The context has.
Knowing how to use AI does not automatically tell the student what to do next. The student has to recognize what changed and decide whether those differences should change how — or whether — AI is used. This situated judgement is a critical workplace skill for students to develop.
Community colleges may be especially well positioned to help students develop that kind of judgment because our students already move between learning and work. According to the Community College Research Center, 67% of community college students work while enrolled. For many, college and work are not separate stages of life. Students may move from a writing course to a manufacturing floor, healthcare setting, office, auto shop, childcare center, or other workplace within the same day.
That movement between school and work is often discussed as a challenge for community college students, and indeed it can be. In the age of AI, I think it may also be one of our educational advantages.
Beyond just using AI
Community colleges are already doing important work to prepare students for an AI-enabled workforce. As AI becomes integrated into occupations well beyond technology fields, students need opportunities to understand these tools, use them effectively, recognize their limitations and consider the ethical implications of their use.
My own research with adult learners in two-year postsecondary writing courses began with a related question: What happens when AI literacy is intentionally taught rather than left for students to figure out on their own?
Students participating in a scaffolded, AI-infused writing curriculum reported increased confidence using AI, clearer understanding of ethical boundaries, and greater recognition of AI’s relevance to future academic and professional writing. In open-ended responses, students imagined using AI for workplace communication, documentation and report writing, brainstorming, technical support and administrative tasks. I found those results encouraging. But I kept coming back to another question.
Recognizing that AI could be useful in future work is not the same as knowing whether it is appropriate for a particular workplace task.
A student may understand the importance of verification, transparency, appropriate use and human responsibility in a college course. But what happens when the audience changes? What if the information becomes sensitive? What if the consequences of an error become greater? What if workplace policy differs from what was permitted in the classroom?
The more I thought about transfer, the more interested I became in what happens when a student cannot simply carry the same AI practice from one setting to another.
Learning to recalibrate
I have begun thinking about this capacity as recalibration: recognizing when a change in context should change how — or whether — we use AI.
That idea builds on a larger shift in how AI competency is being understood. The UNESCO AI Competency Framework for Students emphasizes critical judgment, human responsibility, ethical use, and competencies that can help students adapt to solving problems in new contexts rather than simply learning how particular AI tools work.
But there is an important complication when we think about transfer from college to work. Successful transfer cannot always mean doing in a new environment what worked in the previous one. Sometimes what transfers is the capacity to recognize the difference.
A practice that was appropriate in an English course may need additional safeguards in a workplace. A low-stakes use of AI may require much more verification when the consequences of an error increase. Information that could appropriately be entered into an AI tool in one setting may be protected in another. And an activity that appropriately supports someone’s thinking may become problematic when AI begins substituting for professional judgment the person is expected to exercise.
Sometimes successful transfer may mean deciding to use AI differently — or not at all.
That question shaped my current line of research on situated AI judgment: how students make decisions about AI when important features of a situation change. I am particularly interested in the points where responsible use becomes less about knowing a universal rule and more about reading the situation well.
Change the context and ask again
That research question has also changed how I think about teaching AI. Faculty do not necessarily need an entirely new AI assignment to help students practice this kind of judgment. We can take situations students already encounter and introduce a meaningful change. Give students a situation involving AI and ask them to make a decision. Then change one important variable and ask again.
Suppose AI is used to help draft a routine email confirming an appointment. Would you use it? Now imagine that the email includes confidential patient or client information. Would your answer change? Why?
Or consider a student using AI to brainstorm questions that might be useful when interviewing a hypothetical job candidate. Now change the situation: AI is being asked to evaluate an actual applicant and recommend whom to hire.
The most useful question is not simply, “Can AI do this?” Instead, we can ask students: Would you use AI here? What changed? What would you need to verify? What responsibility still belongs to you?
There may not always be one correct answer. In fact, some of the most useful scenarios may be those in which “yes, but only if…” is a reasonable response. Students can consider whether responsible action means using AI with safeguards, verifying its output, disclosing its use, consulting someone else or deciding not to use it.
The goal is not to make students afraid of AI. It is to give them practice making decisions about it — and those decisions will not end when they leave our classrooms.
An advantage we already have
I hear that need in conversations closer to home. Our employer advisory boards are increasingly discussing AI and asking how students are being prepared to use it effectively and responsibly in the workplace. What employers need is not simply graduates who have used AI before, but people who can make thoughtful decisions about its use as situations and responsibilities change.
That local conversation reflects a broader workforce shift. The World Economic Forum’s “Future of Jobs Report 2025“ identifies AI and big data among the fastest-growing skill areas while also identifying analytical thinking as employers’ most sought-after core skill and emphasizing capabilities such as resilience, flexibility and agility.
Preparing students for an AI-enabled workplace, then, cannot simply mean maximizing their use of AI. It means helping them become professionals who can decide when AI adds value, what safeguards its use requires, and when responsibility should remain with the human.
Community colleges do not need to manufacture a bridge between education and work. Many of our students cross that bridge every day.
They move between classrooms and workplaces, academic expectations and professional ones, low-stakes practice and situations with real consequences. They already know something about adjusting to different contexts. We have an opportunity to make those crossings more visible and educationally purposeful — to help students notice what has changed, determine why it matters and reconsider their use of AI accordingly.
AI readiness is not demonstrated by using AI everywhere it can be used. It is demonstrated by recognizing when the situation has changed enough that our use of AI should change with it.
