The electric utility industry faces two transformations at once. The grid is becoming more digital, data-intensive, and AI-enabled. At the same time, utilities need to develop a new generation of engineers, operators, technicians, planners, and field leaders capable of managing a system facing unprecedented demand and complexity.
T&D World is treating those challenges as connected. The 2026 T&D World Live program highlighted both operationalizing artificial intelligence and building the next-generation workforce. Utilities should do the same by making workforce development part of their AI strategy.
The workforce warning extends beyond the power sector. The Stanford Digital Economy Lab’s August 2026 update found that the employment shortfall for workers ages 22 to 25 in highly AI-exposed occupations widened from 15 percent in the July 2025 data vintage to 19 percent by June 2026. That finding is descriptive, not a causal estimate for utilities. But the underlying issue is familiar: When automation takes over routine analytical work, organizations have to protect the path by which beginners become experts.
Utilities should adopt a digital apprenticeship standard for every major AI-enabled workflow.
The standard would begin by separating repetitive effort from developmental value. AI can summarize inspection data, flag anomalies, help prioritize vegetation management, support load forecasting, and accelerate asset-risk analysis. Some of that work can be automated without sacrificing much developmental value. Other tasks expose developing employees to the patterns and exceptions they will eventually need to recognize independently.
Those tasks should be preserved in a supervised rotation. A junior engineer could review selected AI-flagged asset conditions before seeing the recommended priority. A developing operator could work through abnormal scenarios and explain escalation decisions. A planner could compare an AI-supported forecast against underlying constraints and defend where human judgment should override the model.
T&D World has recently emphasized that utility AI adoption depends on trust, integration, explainability, and human oversight. Oversight becomes stronger when employees have practiced the underlying work themselves.
The standard should define progression. First observe. Then diagnose. Then recommend. Then execute reversible actions. Finally, own a defined decision class independently. Utilities can connect that ladder to existing apprenticeship and qualification systems rather than building a parallel HR program.
This matters because the grid does not forgive shallow competence. When conditions are normal, automation can perform well. Expertise becomes most valuable when conditions depart from the expected patterns, multiple systems interact unexpectedly, or a tradeoff has no clean numerical answer.
Utilities should measure time to independent competence alongside time saved. They should track how many meaningful cases developing workers handle, how often they correctly challenge automated recommendations, and whether the succession bench for critical roles is getting deeper or thinner.
The approach also helps capture the benefit of AI without turning veteran knowledge into a bottleneck. Experienced employees can spend less time on repetitive preparation and more time explaining the cases that encode decades of tacit knowledge. AI can help capture those cases, generate practice scenarios, and organize lessons from past incidents.
That is a more useful productivity story than simple labor reduction.
T&D World recently highlighted the role of mentors and hands-on experience in the career journey of Ameren Illinois leadership. The next generation still needs those ingredients. AI should make them easier to deliver at scale.
The grid of the future will combine more automation with greater operational complexity. Utilities therefore need more than AI-ready infrastructure. They need a repeatable way to develop AI-ready human judgment.
A digital apprenticeship standard would help ensure modernization strengthens both.
