Most organisations frame the AI skills gap as a shortage of data scientists and machine learning engineers. Research consistently shows the real shortage is elsewhere: in the functional AI expertise that bridges domain knowledge and AI capability, in the AI literacy that knowledge workers need to use tools effectively, and in the change leadership that transformation requires. Solving the wrong problem is expensive.

The Wrong Diagnosis

When enterprise leaders talk about the AI skills gap, they typically describe it as a shortage of technical talent: not enough machine learning engineers, data scientists, or AI architects to build and maintain the systems the organisation needs. This framing leads to a predictable set of responses: aggressive hiring for technical roles, investment in academic partnerships, and competition for a limited pool of specialists that drives salaries to levels that most non-technology companies cannot match.

The data suggests this diagnosis, while not entirely wrong, is addressing a secondary constraint rather than the primary one. IDC projects that over 90 per cent of global enterprises will face critical skills shortages by 2026. AI talent demand exceeds supply by approximately 3.2 to 1, with over 1.6 million open AI-related roles globally against roughly 518,000 qualified candidates. The technical talent shortage is real.

But DataCamp's 2026 State of Data and AI Literacy Report, drawing on a survey of over 500 enterprise leaders across the US and UK, found that the biggest capability breakdowns identified by leaders were not in advanced engineering. They were in foundational AI literacy across the workforce, in the ability to evaluate AI outputs critically, and in the capacity to integrate AI tools into actual day-to-day workflows rather than using them as occasional productivity aids.

Most organisations do not lack AI tools. They lack applied workforce fluency. That is a different problem with a different solution.

2x

Organisations that pair AI investment with structured, workforce-wide AI literacy programmes are nearly twice as likely to report significant AI ROI as those that invest in technology without upskilling the people who use it. The return on capability investment is more consistent than the return on technology investment alone. (DataCamp State of Data and AI Literacy, 2026)

The Three Real Gaps

Gap one: Functional AI expertise

The scarcest capability in enterprise AI is not the ability to build models. It is the ability to understand a business function deeply and know what AI can and cannot do within that function, in a way that produces genuine operational insight rather than technically correct but practically useless recommendations.

A machine learning engineer can build a demand forecasting model. What they frequently cannot do is sit with an operations team and understand the specific patterns, exceptions, and business rules that make the difference between a model that works in production and one that produces outputs the operations team does not trust. That bridge, between deep functional knowledge and AI capability, is the capability that most organisations are actually short of.

PwC's AI Jobs Barometer consistently identifies roles at the intersection of domain expertise and AI fluency as the fastest-growing and most competitive in the labour market. AI-exposed roles are evolving 66 per cent faster than other job categories. The average time to fill an AI role has climbed to 68 days, from 42 days in 2023. In regulated sectors such as financial services and healthcare, specialist hiring cycles extend to six to seven months.

Organisations that are closing this gap fastest are not doing so primarily through external hiring. They are investing in the development of their existing functional experts, giving operations professionals, finance teams, commercial managers, and HR business partners the AI fluency they need to become the bridge between their function and the technical teams. This approach produces better outcomes than hiring technical staff and asking them to develop domain knowledge, because domain knowledge is the harder and slower capability to build.

Gap two: Workforce AI literacy

The second gap is much wider and affects every function in the organisation. It is the difference between a workforce that has access to AI tools and one that uses them in a way that changes how work actually gets done.

Gallup's 2026 research found that while half of US employees now use AI at least occasionally, only about one in ten strongly agree that AI has fundamentally changed how work gets done in their organisation. The tools are deployed. The behaviour has not changed. This is a literacy gap, not a tool gap.

The specific capabilities that leaders most frequently identify as missing are not advanced. They include knowing how to prompt AI tools to produce useful outputs rather than generic ones, knowing how to evaluate AI-generated content critically rather than accepting it uncritically, and knowing how to integrate AI assistance into actual workflows rather than treating it as an occasional aid for specific tasks.

These are not skills that require months of training. BCG research found that organisations with formal AI training programmes see 2.3 times faster AI adoption and 67 per cent higher AI ROI compared to those still battling talent gaps. DataCamp's research found that organisations with mature workforce-wide upskilling programmes are nearly twice as likely to report significant AI ROI. The return on investing in foundational AI literacy is better evidenced than the return on most AI technology investments.

Gap three: AI change leadership

The third gap is the least discussed and arguably the most constraining. It is the shortage of leaders who can lead AI-driven transformation rather than just AI deployment.

AI transformation requires a different set of leadership capabilities from AI adoption. It requires the ability to identify where processes need to be redesigned rather than automated, to manage the cultural resistance that comes with significant changes to how work is done, to build trust in AI outputs among teams who are sceptical, and to hold the programme accountable for business outcomes rather than technology activity.

Deloitte's 2026 Global Human Capital Trends research found that while 85 per cent of leaders say building their organisation's ability to adapt at speed is critical, only 7 per cent believe they are actually leading on that front. The gap between recognising the need and having the capability to meet it is where most AI transformation programmes stall.

Why the Technical Hiring Approach Has Limits

The instinct to address the AI skills gap through aggressive hiring for technical talent has several structural problems that are worth naming explicitly.

First, the salary requirements are prohibitive for most non-technology companies. Senior AI engineers command salaries that have increased substantially and that compete with technology sector employers who will always have structural advantages in total compensation. Building the AI capability an enterprise needs through senior technical hires is an expensive strategy that most organisations cannot sustain.

Second, technical talent without domain knowledge produces technically correct but functionally inadequate solutions. The failure mode is common: a technically capable team builds an AI system that the business function does not adopt because it does not match how the work actually gets done. The technical skills were not the constraint. The functional understanding was.

Third, technical expertise attrits quickly. The average tenure of senior ML engineers at non-technology companies is under two years. Replacing a senior AI hire costs roughly 50 to 75 per cent of their annual salary. An AI capability strategy built primarily on external technical hiring is structurally fragile.

What a More Effective Approach Looks Like

The organisations that are closing the AI skills gap most effectively are combining three approaches that address the three real gaps simultaneously.

1. Develop existing functional experts into AI-fluent practitioners. Identify the strongest operations professionals, finance analysts, commercial managers, and HR business partners in the organisation and invest in their AI fluency. The goal is not to turn them into engineers. It is to give them enough AI literacy to lead AI initiatives in their function, evaluate vendor claims critically, and serve as the functional bridge between domain knowledge and technical capability.

2. Build workforce-wide AI literacy as an ongoing capability, not a training event. Role-specific, workflow-integrated learning delivered continuously outperforms scheduled classroom training consistently. The investment does not need to be large. The evidence suggests that structured programmes delivering even a few hours per week of relevant, contextual AI skill development produce measurably better adoption and ROI outcomes than unstructured self-directed learning.

3. Develop AI change leadership at the manager and director level. The manager layer is where AI adoption either compounds or stalls. Investing specifically in managers' ability to lead AI-enabled change, model AI use themselves, create psychological safety for experimentation, and hold teams accountable for behaviour change rather than tool usage is the highest-leverage capability investment most organisations can make.

66%

faster evolution rate of AI-exposed roles compared to other job categories, with the average time to fill climbing from 42 to 68 days in two years. Workers with AI skills command wage premiums of up to 67% over peers. But the scarcest and most valuable capability is not technical expertise alone. It is functional AI fluency: the ability to bridge deep domain knowledge and practical AI capability. (PwC AI Jobs Barometer, 2026)

The Capability Investment That Changes the Return on Everything Else

Enterprise AI investment is substantial and growing. IDC estimates that skills shortages may cost the global economy up to 5.5 trillion dollars by 2026 in product delays, quality issues, missed revenue, and impaired competitiveness. That figure reflects the aggregate cost of deploying AI without the capability to use it well.

The PwC AI Jobs Barometer finding that the most AI-exposed companies achieve 163 per cent productivity growth compared to the least exposed is instructive. Those companies are not simply deploying more AI. They are building the workforce capability to realise what the AI makes possible.

The AI skills gap is a real and significant constraint on enterprise AI value. But solving it by competing for a limited pool of technical talent is both expensive and insufficient. The organisations that will look back on this period as the one where they built a durable AI advantage are those that invested in functional AI fluency across the business, in workforce-wide AI literacy as an ongoing capability, and in the change leadership that turns AI deployment into AI transformation.

That investment is neither expensive nor technically complex. It is primarily a decision about where capability development sits in the organisation's AI strategy, and whether it is treated as a workstream in its own right or an afterthought to the technology investment.

About VE3

VE3 is a global enterprise AI, data, and digital transformation consultancy and Microsoft Solutions Partner. We work with organisations to build the AI capability their transformation programmes require, combining functional domain expertise with technical delivery, workforce AI literacy development, and the change leadership frameworks that make AI investment produce lasting business value. To know more, contact us.