AI use is growing but the skills gap is not going away

The biggest enterprise AI constraint is becoming a people problem

Enterprise AI has a surprisingly human bottleneck. The main source of friction is not access to computing power or its cost. It is whether employees and organizations know how to use AI effectively across real business processes.

Many employees already use AI in their personal lives and have carried that familiarity into the workplace. But most experimentation remains focused on relatively simple, repetitive tasks. Writing drafts, summarizing information, or accelerating routine work can generate useful efficiency gains, but these activities do not automatically transform how a company operates.

The harder question is what happens to an entire workflow when AI becomes part of it. Which tasks should AI perform? Where should a person remain responsible? What information can the system access? Who validates its output? And how should the process change as models and business requirements improve? These are organizational design decisions, not questions individual employees should be expected to solve on their own.

Robinson, whose title and company are not identified in the provided source text, made this distinction clear: “They’re struggling to see exactly how their entire workflow is going to transform. And that’s not really on them.” Robinson placed responsibility on enterprise leaders to integrate AI, develop workflows that can adapt, and identify valuable use cases.

For CEOs and CIOs, this changes the investment equation. Buying more AI tools will not necessarily produce proportional gains. Companies need a people strategy alongside the technology strategy. That means defining valuable use cases, redesigning processes, establishing accountability, and giving employees the training required to work with AI safely and productively.

The objective should also be larger than increasing individual productivity. The real opportunity is to redesign operations so that AI improves speed, quality, decision-making, and eventually the economics of entire processes. That requires coordinated leadership rather than scattered employee experimentation.

Organizations that get this right can move from isolated AI use to repeatable operational capability. The technology is advancing quickly. The executive challenge is to make sure the organization can advance with it.

AI literacy turns technology spending into business value

AI budgets are growing, but larger budgets do not guarantee larger returns. As CIOs become more concerned about uncontrolled spending, the ability of employees to use AI effectively becomes an important part of the ROI equation.

DataCamp found that a larger share of leaders who reported positive returns from AI investments also reported having more mature AI literacy initiatives. The supplied source does not provide percentages, sample sizes, or a study date, so the finding should be treated as an association rather than proof that training alone causes higher ROI.

The practical point is still important. Companies can deploy advanced AI systems and give thousands of employees access, but access does not create capability. Employees need to understand what AI can do, where it can produce unreliable results, how to validate its output, and when human judgment remains necessary. Training also needs to connect these skills to actual workflows rather than teaching AI as an abstract technology.

For executives, this means AI literacy should be part of the investment plan from the beginning. A useful program can cover basic AI concepts, effective tool use, data security, output verification, governance requirements, and role-specific applications. Different teams need different levels of expertise. A finance executive validating AI-supported analysis has different requirements from a software engineer building an AI-enabled product.

Companies should also measure whether training changes business outcomes. Course completion is easy to count but says little about value. More useful measures include adoption in approved workflows, time saved, improvements in output quality, reductions in errors, employee proficiency, and financial returns from specific AI use cases.

This creates a stronger framework for controlling AI spending. Instead of treating technology procurement and workforce development as separate budgets, leadership can evaluate them as parts of the same transformation program. If an AI system cannot be used effectively by the people responsible for producing business results, its technical capabilities have limited economic value.

The goal is not to turn every employee into an AI specialist. It is to create enough AI literacy across the organization that people can use these systems productively, question their outputs when necessary, and identify opportunities for better processes. As enterprise AI investment expands, that capability becomes increasingly important to converting spending into measurable returns.

AI requires a different standard of trust

AI changes a basic assumption employees have learned from decades of using business software. Traditional software usually follows fixed rules: given the same inputs and conditions, it is designed to produce predictable results. Generative AI works differently. It produces outputs based on statistical patterns and probabilities, which means a response can appear convincing while still being incomplete, irrelevant, or factually wrong.

That distinction matters at enterprise scale. Employees cannot assume that an AI-generated answer is correct simply because the system presents it confidently. They need to evaluate two separate questions: Is the output useful for the task, and is the information itself accurate? The level of verification should also increase with the consequences of an error, particularly in areas such as finance, legal work, cybersecurity, healthcare, and regulatory compliance.

Robinson, whose title and company are not identified in the supplied source, described this as a new concept for many workers. As Robinson explained, employees must understand what it means to use software that provides “an answer that needs to be validated, not just for the applicability to the question at hand but even its correctness.”

For executives, this has direct implications for AI training. Teaching employees how to write prompts or operate an AI application is only part of the requirement. Organizations also need employees to understand uncertainty, verify important claims against trusted sources, recognize when specialist review is necessary, and know which decisions should not be delegated to AI without human oversight.

Governance should reflect the same principle. A low-risk internal task may require relatively light validation, while an AI-generated recommendation affecting customers, financial reporting, safety, or regulated decisions can require formal controls, auditability, and clear human accountability. Applying the same level of oversight to every use case can create unnecessary friction; applying too little oversight can create material business risk.

This does not reduce AI’s potential. It clarifies how companies can use it responsibly at scale. AI can accelerate research, analysis, content production, software development, and many other activities. But its value increases when employees understand both its capabilities and its limits.

The organizations that develop this judgment across their workforce will be better positioned to expand AI beyond basic experimentation. The objective is not unconditional trust or constant skepticism. It is calibrated trust: use AI where it performs well, verify what matters, and keep accountability clear.

CIO-HR collaboration can accelerate enterprise AI adoption

AI transformation is not an IT project alone. CIOs can select platforms, establish technical standards, manage security, and define how AI connects with enterprise systems. But widespread adoption depends on people changing how they work. That makes HR a critical partner.

Robinson, whose title and company are not identified in the supplied source, argues that CIOs should establish a clear line of communication with HR leaders. As Robinson put it, “IT leaders may not be fully aware of what’s available from the HR perspective when it comes to learning and development.” That gap matters because many organizations already have training infrastructure, skills frameworks, career-development programs, and change-management capabilities that can support AI adoption.

The division of responsibilities can be clear. IT should define approved technologies, security requirements, data controls, and technical capabilities. HR can identify skills gaps, organize learning programs, support role changes, and measure workforce readiness. Business-unit leaders then connect these capabilities to specific processes and measurable outcomes. Executive sponsorship keeps these efforts aligned with company strategy.

This coordination becomes especially important as AI changes job design. Some tasks will become automated, others will become faster, and new responsibilities will emerge around reviewing AI output, managing exceptions, and supervising AI-enabled processes. Leaders need to understand these changes before deciding whether to retrain employees, redesign roles, recruit new capabilities, or reallocate staff.

Training should therefore be based on roles and business requirements rather than delivered as one generic AI course. Executives need enough knowledge to make investment and governance decisions. Managers need to know how to redesign workflows and evaluate results. Employees using AI directly need practical skills in validation, security, responsible use, and the applications relevant to their jobs.

For the C-suite, there is also a measurement issue. CIOs and HR leaders should agree on outcomes before launching major programs. Useful indicators can include employee proficiency, adoption of approved AI tools, workflow changes, productivity improvements, error rates, time saved, and business value generated. Measuring only training participation reveals little about whether workforce capabilities are actually improving.

The larger objective is to make AI adoption an organizational capability rather than a collection of isolated technology deployments. CIOs understand the systems. HR understands workforce development. Business leaders understand the processes where value must be created. Connecting those perspectives can help companies move faster while maintaining security, accountability, and clear business priorities.

Key executive takeaways

• Workforce readiness is the real AI constraint: Personal AI use does not automatically translate into enterprise capability. Leaders should redesign workflows, define high-value use cases, and build workforce skills alongside technology investments.
• AI literacy directly supports ROI: Access to AI tools is not enough to create business value. Build role-specific training around practical use, validation, security, and governance, then measure improvements in productivity, quality, and financial outcomes.
• AI requires calibrated trust: Probabilistic AI outputs can be useful without always being correct. Match validation and human oversight to business risk, especially for financial, legal, safety, and regulated decisions.
• CIO and HR alignment can accelerate adoption: IT understands the technology while HR brings workforce development capabilities. Connect both functions with business leaders to identify skills gaps, redesign roles, and turn AI deployment into scalable operational change.