Agentic AI could help close the gap between investment in AI and the low returns it offers businesses today. David Lee outlines its potential to future-proof growth and profitability
The disconnect between the efficiency gains promised by artificial intelligence (AI) and its impact on corporate balance sheets is among most significant challenges facing businesses today.
PwC Ireland’s latest CEO survey revealed that 94 percent of chief executives expect AI to be embedded in their workflows within three years. Less than a quarter can demonstrate any meaningful profitability improvements from their investment in AI, however.
This gap demands attention as organisations move beyond AI experimentation.
With close to one-third of Irish CEOs believing their organisation won’t exist in its current form 10 years from now, there is greater pressure to deliver higher returns from AI investment.
Agentic AI—technology capable of autonomous decision-making and actioning—could offer the requisite bridge between personal productivity improvements and enterprise-wide transformation.
The state of AI adoption
AI sentiment around boardroom tables presents a striking paradox. Despite operating in unparalleled macroeconomic conditions, 93 percent of Irish business leaders maintain a remarkably positive outlook on revenue growth, according to PwC’s CEO Survey.
This optimism exists alongside a profound recognition of the need for internal transformation, however.
Close to 30 percent of Irish CEOs do not believe their organisation will exist in its current form within a decade. This creates a strong case for AI investment as business leaders race to reinvent their organisations.
Six-month trends reveal an acceleration in structured AI implementation, with the proportion of Irish organisations kickstarting formal plans and active projects jumping from 50 to 70 percent.
Herein lies the central challenge. While efficiency improvements are widely evidenced, only a quarter of these organisations have translated such gains into profits.
This value leakage—from potential to profit—demands explanation.
Agentic AI to the value gap
If conventional AI has delivered incremental benefits without proportional financial returns, Agentic AI could offer a more compelling proposition. The distinction is not merely technical but fundamental to how value is created and captured.
Agentic AI—systems capable of autonomous decision-making, action-taking and process optimisation—represents a shift from what might be termed “intelligent data manipulation” to “intelligent workflow execution”.
This transition is the difference between personal productivity and enterprise productivity; between automating discrete tasks and reimagining entire processes.
Diverse applications from all areas of the business can be united in their focus on end-to-end processes, rather than isolated tasks. This is precisely the shift needed to bridge the gap between efficiency and profitability.
Strategic implementation framework
Translating Agentic AI’s potential into sustainable financial returns requires a deliberate approach that strikes a balance between innovation and pragmatism. The following framework offers a pathway.
The progression from conventional to Agentic AI implementation is evolutionary rather than revolutionary.
The most successful organisations establish proof points through targeted deployments before attempting wholesale business model reinvention.
This approach creates the reference experiences necessary to build internal confidence and stakeholder support.
Successful and sustained AI adoption must also address obstacles simultaneously. A sequential approach—solving technical challenges before addressing governance concerns, for example—invariably creates impediments to scale.
The most effective organisations pursue parallel workstreams that address technology implementation, organisational capability building, governance development, stakeholder engagement, cybersecurity and security enhancement.
Particular attention should be paid to the behavioural change requirements. The adoption curve for AI follows predictable patterns—early enthusiasts, the pragmatic majority and reluctant laggards.
Effective adoption strategies account for these different constituencies, rather than designing exclusively for the enthusiasts.
The behavioural shifts required to support Agentic AI extend beyond initial adoption to continuous learning as capabilities evolve. This differs from the “train once” deployment models of traditional technology implementations.
Implementation must also proceed at a pace that maintains trust across all stakeholder groups.
Trust, once compromised, requires disproportionate effort to restore—a calculation that justifies measured progress over hasty deployment.
Balancing innovation and pragmatism
The value gap between AI’s promised benefits and its profit delivery represents the central challenge for business leaders navigating the current wave of technological disruption.
With nearly a third of Irish CEOs questioning their organisation’s future in its current form, the imperative to bridge this gap has never been more acute.
Agentic AI offers a pathway from incremental improvement to fundamental transformation by shifting focus from isolated task automation to orchestrated process reimagination.
Organisations demonstrating measurable financial returns have moved beyond the “faster horses” mindset to rethink how work itself should be structured and executed.
Yet, technology alone cannot close the value gap. Successful implementation requires simultaneous attention to business case development, organisational capability building, governance structures, stakeholder trust and security considerations.
The most effective approaches strike a balance between innovation ambition and implementation pragmatism, building reference experiences before attempting wholesale business model reinvention.
The most valuable lesson from early adopters is perhaps counterintuitive: the strongest financial returns often come, not from cost reduction through displacement, but from capacity expansion through augmentation.
As organisation’s progress from experimentation to enterprise adoption, they would do well to remember that AI is not just a “new tool”. Rather, it represents a fundamental shift in how work is conceived and executed.
Those who approach it merely as a means to do existing things more efficiently will find themselves with faster horses in an age that demands flying cars.
David Lee is Chief Technology Officer at PwC Ireland