James Madison University has opened a door that too many AI workforce plans leave closed. Its Bridge2AI program, announced Oct. 19, offers adult learners from underserved and rural communities a structured route into AI-related careers. The program combines months of training with mentorship and a potential 10-week paid internship.
There is an important detail in the fine print: the internship is not guaranteed. It would take place at CGI in Lebanon, Virginia, and participants generally need to live within commuting distance. The program recognizes that childcare, transportation, and access to equipment can determine whether a motivated adult gets a realistic chance.
Virginia employers should pay attention. The challenge facing someone who works a shift in Staunton or cares for a parent outside Waynesboro is rarely a lack of enthusiasm for technology. That person needs a schedule that works, a mentor who can answer questions, and evidence that new skills will lead somewhere.
I study the psychology of workplace AI adoption. A pattern I encounter in organizations is straightforward: enthusiasm rises during demonstrations and falls when employees return to jobs where nobody has adjusted their responsibilities or made room for practice. Employers report wanting AI-capable workers, yet the entry-level tasks that once taught people judgment are increasingly tempting to automate.
A credible workforce bridge must protect that learning process.
The U.S. Department of Labor’s registered apprenticeship model includes paid work, structured on-the-job learning, experienced mentors, related instruction, and recognized credentials. Local businesses do not need to copy every feature of a formal registered apprenticeship to learn from that combination. They can pay a trainee to prepare draft reports with AI, then require the trainee to catch incorrect numbers, trace a claim to original records, explain an exception, and discuss the decision with a customer.
Such assignments teach skills that survive the next software update. Someone who merely learns to produce an attractive AI summary has limited leverage in the workplace. Someone who can recognize an invented fact, explain its consequences, and fix the workflow has a more durable role.
The Shenandoah Valley also needs its own practical placement network. A paid internship in Southwest Virginia serves eligible applicants near Lebanon. It does not solve commuting barriers for everyone elsewhere in the Commonwealth. Regional hospitals, manufacturers, local governments, and small professional firms could each sponsor a few supervised, paid placements tied to real business processes. Community colleges and universities could coordinate common competency tests so employers understand what trainees can actually do.
Measure those placements differently from typical training publicity. Report how many applicants could participate despite childcare or transportation demands, how many completed supervised assignments, how quickly they learned to identify AI errors, and how many moved into paid work. Count the cost of mentoring alongside output improvements. That would show whether employers are creating capability or simply shifting training costs to workers.
Bridge2AI provides one promising design. Virginia should build on its strengths without confusing an available educational pathway with a guaranteed job. The next measure of success belongs on a paycheck and in a manager’s assessment of independent judgment, not on a certificate alone.
Gleb Tsipursky, PhD, 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)
What better way than an election to celebrate the 250th anniversary of the signing of the Declaration of Independence.
Across the world last week, democracy was victorious. So many people voted in presidential elections that their will could not be denied.
AFP will break down local, Virginia and national races in the battle for the control of Congress Tuesday night.
