PwC's 2026 CEO survey finds that only 17% of Irish chief executives say AI has delivered increased revenues in the past twelve months — compared to 29% of their global peers. Almost 75% of AI's economic value is being captured by just one-fifth of companies. The gap is not about technology. It is about approach.
PwC published its 2026 CEO Survey this week. The Ireland-specific findings on AI profitability deserve serious attention from leadership teams across the country — not because they are surprising, but because they confirm a pattern that is already visible in how most Irish organisations are engaging with AI.
Seventeen percent of Irish chief executives say AI has delivered increased revenues in the past twelve months. Their global peers report 29%. On cost reduction, 24% of Irish respondents say AI has helped versus 26% globally. The gap is not dramatic in absolute terms. Its significance is in what it reveals about the nature of AI adoption in Ireland — and what distinguishes the organisations that are capturing value from those that are not.
The pilot mode trap
The PwC survey describes most organisations globally as stuck in "pilot mode." Only 17% of companies are capturing the substantial majority of AI's economic value — nearly 75% of total gains. The other 83% remain in exploration or pilot phase, running multiple AI experiments that have not yet translated into measurable financial returns.
The pattern in Ireland is more pronounced than the global average. Only 8% of Irish CEOs report AI application across a range of business areas, compared to higher proportions among global peers. This suggests that Ireland's AI adoption is not just slower but more fragmented — concentrated in isolated pilots rather than embedded in operational processes.
This is not a technology problem. The tools available to Irish organisations are identical to those available to global peers. The constraint is not access to AI — it is how organisations approach deployment.
What the leaders do that others do not
The PwC analysis identifies a consistent set of behaviours that distinguish the organisations capturing AI value from those in permanent pilot mode.
They point AI at growth, not just cost reduction. The majority of AI pilots in Ireland are focused on efficiency: automating tasks, reducing processing time, cutting administrative overhead. These are legitimate targets, but they are the easiest cases to make and the hardest to scale. The organisations producing measurable revenue impact are deploying AI in customer-facing processes, product development, and market analysis — places where the value created shows up in revenue lines, not just cost lines.
They do the diagnostic first. The organisations that report genuine AI returns have, consistently, taken time to understand where their actual leverage points are before deploying tools. This sounds obvious. In practice, most Irish organisations skip it. Licences are purchased, tools are deployed, and measurement is attempted after the fact. The result is the PwC finding: AI is used, but the impact is dispersed across low-value tasks rather than concentrated where it matters.
They build governance that makes AI scalable. The difference between a pilot and a deployed system is governance: documented processes, accountability structures, quality controls, and feedback mechanisms. Pilots fail to become operations not because the AI does not work, but because the organisation has not built the infrastructure to run AI at scale reliably. Governance is not a compliance overhead — it is the mechanism that converts a tool into a capability.
They measure the right things. Organisations in pilot mode typically measure AI adoption (how many people are using the tool, how frequently) rather than AI impact (what changed in the business because of it). The leaders measure outcomes: revenue generated, decisions improved, time reallocated from low-value to high-value work. You cannot manage what you do not measure, and most Irish AI programmes are measuring the wrong things.
The structural challenge for Irish organisations
There is an honest structural challenge for Irish businesses that is worth naming. Many of the organisations driving global AI returns are large enterprises with dedicated AI teams, significant data infrastructure, and the capital to absorb multiple failed experiments before finding what works. Irish SMEs and mid-sized organisations are working with different resource profiles.
But resource constraints are not the primary explanation for the 17% versus 29% gap. The PwC data suggests that the determinant of success is less about scale than about approach — specifically, whether the organisation treats AI as a strategic question that starts with a diagnostic or as a technology procurement decision that starts with vendor selection.
The organisations that will move from 17% to 29% are not the ones that buy more AI tools. They are the ones that make a more deliberate decision about which problems AI should actually solve, and then build the governance to make those solutions stick.
What to do if you are in the 83%
If your organisation has invested in AI tools without yet producing measurable returns, the question is not whether to continue but where to focus. Three diagnostics are worth running:
Where is the actual leverage in our workflows? Not where AI could theoretically be useful — where would AI meaningfully change the output or quality of work your organisation produces? This requires examining actual working patterns, not hypothetical use cases.
What are we measuring? If your AI metrics are usage-based rather than outcome-based, you are measuring activity rather than value. Define what a successful AI deployment looks like in revenue or cost terms, then measure against that.
What governance do our AI deployments currently have? Documented processes, named accountability, quality controls, incident reporting — these are not bureaucratic additions. They are the infrastructure that lets AI operate reliably at scale rather than producing inconsistent outputs in isolated pilots.
The PwC findings are a useful market check, not a reason for alarm. The organisations that capture AI value are not doing something magical — they are being more deliberate about diagnosis, governance and measurement. That is replicable. If you want to assess where your organisation sits and what would move you from pilot mode to measurable return, that is a conversation worth having before the gap widens further. At Acuity AI Advisory, it is how we start every engagement.