Gerard Hayes, a partner at BearingPoint Ireland, discusses findings from the company’s ‘Scaling AI for Measurable Impact’ report.

As artificial intelligence continues to evolve at a dizzying pace, businesses and other organisations are attempting to grasp the technology’s implications and potential.

If adopting AI now seems non-negotiable, organisations must also consider how to ingratiate it with their existing processes and structures – or how it might dictate them moving forward.

A recent study by consultancy firm BearingPoint has found that the scaling of AI marks the next big hurdle around the technology for those organisations that employ it.

This is more true in Ireland than globally, according to the data, which shows that just 4pc of Irish organisations that have implemented AI say they have completely achieved their scaling objectives, compared to 13pc elsewhere.

Measurable impact

BearingPoint conducted a global survey of 1,050 C-suite executives and senior leaders across 13 countries to assess the correlation between AI investment and measurable value for public and private sector organisations in Europe, the US and China.

The report, titled ‘Scaling AI for Measurable Impact’, finds that although organisational AI maturity is progressing, deep enterprise integration remains uncommon.

“The conversation is moving from whether AI can create value to whether organisations can scale that value,” notes Gerard Hayes, a partner at BearingPoint Ireland.

“I think the report shows that we’ve reached an important point in enterprise AI adoption.”

The research warns of an emerging “AI productivity trap” that companies need to consider. While the tech can make processes faster, automate routine activities and free up employee capacity, the next question becomes how to capitalise on these potential gains.

Deploying more technology does not automatically translate into stronger financial performance, the report argues. Rather, sustainable value “depends on whether organisations can connect AI investment with financial accountability, trusted data, scalable architecture, workforce adaptation, governance and measurable outcomes”.

By extension, organisations risk missing the chance of improving growth, lowering costs or increasing margins if traditional roles, workforce structures, operating models or resource allocation remain unchanged.

If technology creates capacity, argues the report,  then management decisions determine whether that capacity becomes measurable value.

“Many organisations have become very good at launching pilots and quickly demonstrating that an AI use case can work. That’s important progress,” says Hayes.

“The much harder challenge is moving from a successful pilot to something that can be used securely, responsibly and consistently across the enterprise.

“At that point, questions around governance, policies, regulatory compliance, data ownership and accountability become much more significant.

“If those foundations haven’t been considered from the outset, organisations can find that a pilot which was relatively quick to launch takes much longer to expand.”

According to Hayes, organisations face a considerable challenge in balancing speed and agility with structure and rigour when it comes to AI adoption and scaling.

“They need the speed and flexibility to experiment and learn, but also the discipline to put the foundations required for scale in place. Moving too quickly without those foundations creates risk and rework.”

Moving too slowly, of course, represents another type of risk if it allows competitors to learn faster, build organisational capability earlier or better compound the value they are achieving, he adds.

“We also need to recognise that none of this is easy. The pace of change in the AI market is extraordinary. Capabilities are accelerating, costs are fluctuating and new solutions are emerging constantly, while the longer-term direction of the market remains uncertain.”

The Irish outlook

Hayes, who has been with BearingPoint since 2008, works “at the intersection of business and technology”.

He has spent a large part of his career focused on major transformation programmes, particularly within Ireland’s public sector, building “first-hand” experience of complex transformation challenges and an understanding of what makes change successful – and “equally importantly”, what can get in its way.

Viewing the report from an Irish perspective, he notes a trend of strong activity and ambition around AI in this country facing challenges in translation to enterprise-wide scale, pointing out that while 63pc of Irish organisations surveyed say they have implemented limited AI projects – more than the global rate of 54pc – only 6pc in Ireland describe AI as fully integrated across their organisation, compared with 11pc globally.

Especially interesting, in his view, is an indicator that Irish organisations foresee agentic AI as a future priority to a greater extent than the global rate – 43pc compared to 26pc – and that this can be regarded as an opportunity to turn Irish appetite for innovation into organisational capability for deploying AI at scale, leading to measurable value.

Additionally, Hayes highlights findings that 80pc of Irish organisations are conducting ongoing skills-gap assessments as they prepare for AI transformation, reflecting, in his view, a crucial understanding that workforce requirements are set to keep evolving.

For him, a company’s AI strategy and workforce strategy must go hand-in-hand.

“As AI becomes embedded into business processes, and particularly as we move towards more agentic forms of AI, organisations need to look beyond individual tasks and think about roles end-to-end.

“From my experience of transformation, organisations have always needed to anticipate how technology will change their workforce, and good workforce planning is inherently continuous.”

Pace of change

BearingPoint’s report highlights a number of fundamental conclusions when it comes to the challenges of scaling AI for organisational impact.

It states that: AI should be managed “as an enterprise value portfolio” for a holistic approach; organisations need to convert AI-created capacity into value beyond simple headcount adjustments; workforce planning needs to “become part of the AI roadmap”; agentic ambition is moving faster than organisational readiness; and AI leaders should adhere to “execution discipline” in order to distinguish themselves.

In Hayes’s view, AI integration and scaling presents some of the same organisational challenges as previous transformation cycles, including figuring out how roles will evolve, which capabilities will be needed, and where new technology can augment or automate work.

However, for him, the evolving age of AI is defined by one characteristic above and beyond previous technological leaps: speed.

“I think the real opportunity for organisations is not simply using AI to make individual development activities faster, but using it to shorten the journey from a business problem or idea through to a working solution and ultimately measurable value.”

Something that “feels different with AI”, he says, is the pace at which the “underlying assumptions” informing workforce planning and strategy are changing.

“The planning horizon hasn’t disappeared, but the cycle between planning, learning and adapting is becoming much shorter.

“You still need a clear view of the workforce you are trying to build, but you also need the agility to continually challenge the assumptions behind it as the technology evolves.”

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.