Beyond upskilling, start redesigning work: why an AI training plan is not enough

In Energy, Infrastructure, Construction, and Power and Utilities, leaders are no strangers to workforce transformation. Digitalization, safety standards, and automation have already reshaped roles and skills.

And yet, when it comes to AI, many organizations are running into a familiar but uncomfortable reality: the technology is moving faster than the workforce system designed to absorb it.

AI is not just another “productivity tool.” It changes how work is done, how decisions are made, and how accountability is distributed. The real question for leadership is no longer whether AI will impact the workforce, but whether the organization is deliberately shaping that impact or letting it emerge by accident.

For many organizations, early AI adoption has focused on tools: copilots, chatbots, and automation features layered onto existing processes. The implicit assumption is that jobs stay largely the same, just faster. That assumption is already breaking down.

As AI systems move from assisting humans to executing parts of workflows and influencing decisions, work is being redesigned in real time. As AI is embedded into daily operations, the underlying tasks within a role are affected in very different ways. Some become AI‑augmented or fully automated, while others demand more human judgment, oversight, and accountability than before.

What changes is not the job title, but the architecture of work. This creates structural tensions:

• Roles that were designed for execution now require judgment
• Managers are expected to lead hybrid teams of humans and AI without clear guidance
• Capacity freed up by AI is captured as efficiency, rather than reinvested into higher‑value work or employee well-being

In asset‑intensive environments, this matters. Poorly designed work doesn’t just frustrate employees, it creates operational risk. When AI is introduced without redesigning roles and decision boundaries, organizations see operational errors caused by unclear handoffs between humans and AI, as well as safety risks when automated outputs are followed without sufficient oversight. Consider a utility company using AI to prioritize asset inspections. If teams assume AI recommendations are always correct, critical equipment may be overlooked or maintenance resources misallocated, creating reliability and safety risks despite the technology being technically sound.

When organizations discuss AI workforce risk, the conversation often gravitates toward employee resistance. However, reports and surveys show that AI adoption is not the challenge. Employees are already using AI, and often even from outside the company’s policies.

In practice, the more immediate risk is far more subtle: ambiguity.

• Which decisions can be informed or executed by AI, and which must remain human‑owned?
• Where is speed prioritized, and where is caution non‑negotiable?
• Which skills actually matter in an AI‑augmented role? And which no longer do?

While the first two questions touch on AI governance and accountability, they have direct implications for workforce design, leadership responsibilities, and capability development.

Without redesigning the surrounding system, organizations end up with high expectations for productivity gains, but no structural changes to actually realize them. This ambiguity is particularly challenging for mid‑level managers and team leads. They are expected to encourage AI adoption, manage performance, and safeguard quality, often without a shared narrative or skill framework to lean on. AI fluency, in this context, is not about prompting techniques. It is about judgment, coordination, and accountability in hybrid human‑AI systems.

Organizations that move beyond isolated AI experiments share a common pattern: they recognize that workforce transformation in the age of AI cannot be solved through training alone.

They make deliberate leadership choices about:

• How work is redesigned, not just which tools are deployed
• What is expected from leaders managing human‑AI collaboration
• Which capabilities matter most, including judgment, orchestration, and ethical decision‑making
• How AI connects to performance, quality, and well-being, not just cost reduction

The organizations that succeed are not those that “upskill the fastest,” but those that weave leadership, work design, and capability development into a coherent system.

If any of this sounds familiar, you are not alone. Leaders across Energy, Infrastructure, Construction and Utilities are navigating the same tension: AI ambition is rising, but the workforce system has not caught up yet.

This is exactly what the AI Workforce module in KPMG’s AI Excellence Program is designed for. The program brings together leaders in a cohort‑based setting to tackle AI workforce transformation where it actually happens: in roles, teams, and leadership behaviors. Participants move beyond abstract discussions to explore how work can be redesigned, how skills priorities shift, and how leaders can confidently manage human‑AI collaboration in regulated, asset‑intensive environments.

If you recognize the gap between AI potential and workforce readiness, this is your next step. More info coming soon.