Carolinas Colleges Need An AI Apprenticeship Test

By GLEB TSIPURSKY, PHD

Students across the Carolinas are getting more chances to learn artificial intelligence. The next test is whether employers and colleges preserve the work that turns beginners into professionals.

At UNC Charlotte, a recent AI Summit for Smarter Learning brought together more than 400 faculty, staff and students around AI in teaching, learning and workforce readiness. North Carolina has also approved its first 43 Workforce Pell training programs, adding another route into short-term career preparation.

Those investments matter because the labor market is changing underneath them. Stanford Digital Economy Lab’s August employment update found that employment among U.S. workers ages 22 to 25 in highly AI-exposed occupations was about 19 percent below where it would have been if it had kept pace with similarly aged workers in less-exposed occupations. The comparable gap was 15 percent in the July 2025 data vintage. The adjustment appears mainly through reduced hiring, especially where AI use tends to automate human tasks.

That should change how colleges define AI readiness.

A student who learns to prompt a model, summarize material or automate a routine analysis has learned something useful. A student who learns to catch a confident error, explain an exception, test an output against reality and defend a recommendation has learned the harder part of professional work.

The problem is that entry-level jobs have traditionally bundled those lessons together. Beginners prepared drafts, cleaned data, assembled research and handled routine client work. Experienced colleagues reviewed the result, corrected mistakes and gradually handed over harder decisions. If AI takes away the preparation work and organizations simply hire fewer beginners, the career ladder loses its first rung.

Carolinas colleges should respond with an AI apprenticeship test for career programs. Every program that teaches substantial AI use should answer three questions before calling students job-ready.

First, what decisions will a new employee own after AI handles the routine preparation? Programs should build assignments around those decisions, including verification, escalation and communication.

Second, where will students practice exceptions? Real expertise grows when the standard answer stops working. Courses should require students to diagnose flawed outputs, conflicting evidence and unusual cases rather than reward smooth automation alone.

Third, who will review their judgment? Employer partnerships should include named mentors, feedback cycles and progressively harder responsibilities. The metric should be time to independent competence, not simply time saved on a task.

This approach fits the region’s existing workforce strategy. North Carolina already treats apprenticeship and work-based learning as routes into high-demand fields. The next step is to extend that logic into AI-exposed office, technical and professional work.

Employers benefit too. A company that automates routine preparation but still develops junior talent gets the productivity gain without starving its future pipeline of experienced managers and specialists. It also creates a clearer answer to the question students increasingly ask: if AI does the beginner work, how do I become an expert?

The answer should be visible in the curriculum and in the workplace. Use AI to shorten the path from novice to trusted professional. Do not shorten the career ladder itself.

Gleb Tsipursky, PhD, is a behavioral scientist, CEO of Disaster Avoidance Experts, and author of The Psychology of AI Adoption at Work: From Resistance to Results (Georgetown University Press, 2026). https://disasteravoidanceexperts.com/aibook