82% of Irish businesses report critical skills gaps undermining productivity. AI and digital skills are among the hardest to fill. But the answer is not always to hire — and for most Irish SMEs, it cannot be. Here is what actually works.
IBEC's most recent data is unambiguous: 82% of Irish businesses are grappling with critical skills gaps that are actively undermining productivity, innovation, and competitiveness. AI and digital skills are consistently cited among the most acute shortfalls. The ERF's characterisation of Ireland as facing a "skills wall" — not a skills gap, but a wall — is not hyperbole.
The broader picture is simultaneously striking and counterintuitive. Ireland has one of the highest concentrations of global technology companies in Europe. AI is a stated national priority with a 90-action Digital Ireland strategy behind it. And yet 83% of employers cannot find candidates with the skills they need — the highest level of talent shortage in 20 years.
Understanding why this is happening is necessary before deciding what to do about it.
Why the shortage is structural, not cyclical
The skills shortage in Ireland is not a temporary tightening of the labour market that will resolve as supply catches up with demand. It is structural, for three distinct reasons.
The nature of AI work is changing faster than education and training can adapt. The AI skills that were in demand two years ago — machine learning engineering, data science, Python development — are partially being automated by the same AI tools that require those skills. The skills in demand today include AI prompt engineering, AI governance, AI-augmented process design, and the ability to supervise and evaluate AI outputs. These are not subjects that universities and professional development programmes have had time to build curricula around. The formal education pipeline is at least three years behind the market.
Ireland's AI demand is three times the European average. 11% of Irish job postings mention AI, compared to approximately 4% in both the EU and the US. This is not because Irish businesses are uniquely innovative. It is because Ireland's economy is concentrated in the sectors — technology, financial services, professional services — where AI integration is happening fastest. The demand signal is proportionally stronger here than almost anywhere else.
The entry-level pipeline has been disrupted simultaneously. The junior workforce that would have spent two to three years in structured, repetitive roles building up tacit knowledge and technical confidence has seen those roles contract sharply. The traditional path from entry-level to mid-level, through which organisations developed their own capability, has been compressed. The mid-level talent shortage of 2028 is being created now.
What does not work
Before describing what does work, it is worth being honest about what does not.
Competing on salary for the small pool of available AI talent is viable for large enterprises with deep pockets. It is not viable for Irish SMEs. The salary premium for genuine AI expertise — particularly in machine learning engineering, AI strategy, and AI governance — is substantial, and the candidates available at that level in the Irish market are few and highly sought after.
Expecting the existing team to upskill through informal learning — giving people access to online courses and hoping capability develops — works for motivated individuals but does not produce organisational capability at scale. The research consistently shows that informal upskilling without structured application, management support, and time allocation produces modest and fragile results.
Hiring generalists who claim broad AI expertise is a particular risk in the current market. The consulting ecosystem has expanded rapidly with practitioners offering "AI strategy" services whose actual competence range from genuinely expert to largely theoretical. The absence of recognised professional standards for AI consultants makes it difficult to evaluate credentials. The temptation to hire or engage quickly — given the urgency many organisations feel — makes due diligence on AI advisory quality important and often neglected.
What does work
Targeted, role-specific upskilling with structured application. The evidence on what produces lasting capability uplift in organisations is consistent: learning that is connected to real work, applied within weeks of the training, supported by managers who have also been trained, and reinforced through peer practice. Generic AI awareness training produces awareness. Role-specific AI tool training connected to actual work processes produces capability.
For most Irish organisations, this means identifying the three to five specific AI applications with the highest near-term value in their operations, training relevant staff specifically on those applications, and building in a structured period of supervised application before the training investment is considered complete.
Building internal AI champions, not AI departments. Large organisations can afford dedicated AI teams. Most Irish SMEs cannot and should not. The more scalable model is identifying two or three people in existing roles who have both the interest and the aptitude for AI work, investing in their capability more deeply, and using them as internal resources who support their colleagues' AI adoption while maintaining their existing functional roles.
This model works best when the AI champions are in operational roles — not in IT — and when they have explicit organisational support including time allocation, access to external learning resources, and management-level visibility.
Using external advisory capacity strategically. The case for external AI advisory support is strongest where the need is for specific expertise on a defined question — a regulatory compliance assessment, a use case evaluation, a governance framework — rather than for ongoing general AI guidance. External advisors with genuine sector expertise and a vendor-neutral position can accelerate decisions that would otherwise take months to work through internally.
The distinction worth drawing is between advisory support that builds internal capability and consulting engagements that create dependency. The former is an investment. The latter is a recurring cost that does not reduce over time.
Engaging the government support infrastructure. Enterprise Ireland's AI Adoption Roadmaps, CeADAR's industry engagement programmes, and the Digital Ireland skills campaign are all designed to reduce the cost and complexity of AI capability building for SMEs. They are underutilised — partly because most SMEs do not know they exist, and partly because engaging with them requires some initiative that organisations under operational pressure often cannot find.
The capability question is a leadership question
The organisations that are building AI capability effectively in Ireland share one consistent characteristic: the decision to invest in capability is owned at leadership level, not delegated to IT or HR.
AI capability building requires decisions about what work the organisation wants AI to augment, what skills are consequently required, what the investment in developing those skills is worth, and how progress will be measured. These are strategic questions. When they are delegated downward, they typically do not get answered — they get managed around the edges while the skills wall remains in place.
The 82% of Irish businesses reporting critical skills gaps are not all making the same mistake. But the single most common factor in organisations that have not made progress on AI capability is that leadership has not made the capability question a leadership priority. Everything else — tool selection, training providers, hiring plans — follows from that decision being made.
The talent is scarce. The wall is real. But for most Irish organisations, the primary constraint is not the labour market. It is the clarity of internal intent.