Opinion: Colorado should turn artificial intelligence skills into apprenticeships

Colorado is getting better at seeing the skills employers want in real time. An August report from The Colorado Sun described how the state’s revamped Connecting Colorado job platform links job seekers with employers and tracks which skills appear in postings. State officials counted 4,574 job postings requesting artificial intelligence as a skill.

That is useful information. The harder question is: What happens after a young worker gets hired into an AI-enabled job?

The Stanford Digital Economy Lab’s August payroll update found that employment among U.S. workers ages 22-25 in highly AI-exposed occupations is about 19% below where it would be if it had kept pace with similarly aged workers in less-exposed occupations. The adjustment appears mainly through reduced hiring. Experienced workers show no comparable gap.

Employers have a straightforward reason to use AI on entry-level work. Software can prepare a first draft, summarize records, classify routine requests or build a preliminary analysis in minutes. The savings are real. The risk arrives when the company treats the disappearing preparation task as evidence that the junior role itself has disappeared.

Those starter assignments once performed a second function. They gave new employees repeated, lower-risk chances to learn the work. A supervisor could see whether the analyst spotted an odd number, whether the coordinator understood an exception, whether the associate asked the right question before sending a recommendation to a client. Repetition created judgment.

Colorado should replace that accidental apprenticeship with a deliberate one. Every AI-enabled entry job should contain a sequence of supervised competency blocks. In the first block, the employee audits AI output against source material and documents errors. In the second, the employee handles exceptions that fall outside the standard process. In the third, the employee explains an AI-assisted recommendation to a customer, colleague or manager. In the final block, the employee completes an end-to-end assignment while an experienced reviewer checks only the highest-risk points.

The state already uses work-based learning to connect education with employer needs. Colorado’s Department of Labor and Employment is actively recruiting business hosts for its 2026 K-12 Teacher Externship Program, designed to bridge education and industry in high-growth STEM occupations. The same practical logic belongs inside AI-enabled workplaces: learn the real work, under real supervision, against clear standards.

Large employers can build these pathways themselves. Smaller firms need a shared structure. Colorado’s workforce centers, community colleges and industry groups can help define portable competency templates that employers adapt to finance, health services, construction administration, professional services, technology and other local sectors. The state does not need one universal AI curriculum. It needs a common architecture for turning AI-assisted work into demonstrable human capability.

That architecture should change what employers measure. Hours saved and output volume still matter. Add time to independent competence. Track how quickly a new employee can catch a model error without prompting, resolve an unusual case, explain the evidence behind a recommendation and know when to escalate. Those measures tell a company whether AI is strengthening its future workforce or merely lowering this quarter’s labor cost.

The choice matters because Colorado’s economy depends on a deep bench of people who can move from technical fluency to sound judgment. AI can shorten that path. It can give beginners more examples, faster feedback and earlier exposure to difficult cases, provided experienced workers remain responsible for coaching and review.

Colorado already knows how to connect skills data, employers and work-based learning. The next step is to make apprenticeship part of AI adoption itself. Every routine task automated should create room for a harder supervised task. Every hour saved should move a junior worker closer to independent competence.

That is how Colorado can turn growing demand for AI skills into a stronger career ladder.

Gleb Tsipursky, of Columbus, Ohio, is a behavioral scientist and the author of the peer-reviewed book, “The Psychology of AI Adoption at Work: From Resistance to Results.”

The Colorado Sun is a nonpartisan news organization, and the opinions of columnists and editorial writers do not reflect the opinions of the newsroom. Read our ethics policy for more on The Sun’s opinion policy. Learn how to submit a column. Reach the opinion editor at opinion@coloradosun.com.

Follow Colorado Sun Opinion on Facebook.