Artificial intelligence is reshaping how organisations operate - faster than some leadership behaviours are evolving to keep pace.
KPMG’s Colm O’Neill brings a dual perspective to this shift, shaped by both his own experience leading a business itself undergoing AI-drive change, and by what he sees working with clients and the wider market.
Artificial intelligence is reshaping how organisations operate - faster than leadership behaviours are evolving to keep pace. Over the past 18 months, AI has moved from experimentation to embedded practice.
Our latest CEO Outlook survey found that nine in ten Irish CEOs (89%) believe AI will deliver competitive advantage in the coming months, with 63% identifying it as a top investment priority.
In many cases, this shift to embedding AI has happened organically, led by teams and tools rather than by executives. The result is a growing gap between where AI is active and where leadership believes transformation is happening.
That disparity is concerning. Because while AI is often framed as a technology initiative, the most significant leadership challenges are often behavioural and organisational.
Leaders are being asked to guide their organisations through a transition where the impact is vast, the timelines are uncertain, and the expectations from investors, boards, and employees continue to rise.
In our experience we clients, we are hearing a move beyond the initial “fear of missing out.” Most organisations have already deployed AI tools - whether Microsoft Copilot or domain‑specific LLMs - and have experimented across various functions.
But now a new dynamic has emerged: a fear of messing up.
Leaders are not only excited by what AI can do; they are increasingly concerned about the risks that might emerge from incorrect use, data exposure, or unintended outcomes. Decision making can get more complex, timelines slide, and stagnation becomes a risk.
At the same time, stakeholders have become more assertive. Companies have spent the past year announcing pilots and deployments, but investors are now asking:
This tension - between the pressure to innovate and the pressure to deliver evidence - is shaping leadership conversations across sectors.
AI is no longer a contained technology project. It is becoming part of how work gets done, often before senior management or business owners have fully aligned on expectations, behaviours, or governance.
That shift elevates the leadership challenge: not just managing technology, but steering an organisation whose processes, decisions, and responsibilities are being continually reshaped by AI.
AI represents both an opportunity and an exposure across every dimension of the organisation. Leaders know AI will change everything - they just don’t know exactly how or when.
That uncertainty makes it almost impossible to define the “total prize” of AI, but it doesn’t remove the urgency to act.
AI systems rarely fail on their own - behaviour does. Over‑reliance on AI outputs, lack of challenge, unclear accountability, and misplaced trust create exposure long before a technical issue emerges.
Leadership mindset, curiosity, and judgement matter more than any policy. Are you using AI yourself? If not, why not? And if so, where?
AI enables automation, but its real value lies in how it reshapes processes, roles, and decision-making. Small Language Models (SLMs), for example, allow leaders to solve highly specific problems with known datasets - offering precision, reassurance, and strong foundations for impact.
These targeted applications help organisations transform cost bases and uncover productivity opportunities grounded in trusted data.
Every organisation - regardless of size - is being reshaped by AI. Smaller organisations often have less margin for error and may play critical roles in broader ecosystems. Thinking you are “under the radar” is increasingly a leadership blind spot.
At its core, this is about how people think, decide, and act.
Leadership actions in this space are practical, not theoretical.
Responsibility must remain with people, not tools. Management should explicitly reinforce this, while demonstrating the behaviours they expect to see; thoughtful challenge, informed curiosity, and ownership of decisions.
Employees take cues from how leaders themselves use AI.
Leaders need to move from pilots to purpose. SLMs offer a powerful route to high‑confidence solutions: finite data, controlled environments, and precise outcomes.
They help organisations move from experimentation to measurable impact - particularly in areas such as cost transformation and productivity.
Fragmented experimentation creates fragmented leadership outcomes. Management should shift from isolated pilots toward enterprise‑wide principles, anchored in clarity about what good AI behaviour looks like.
Consistency builds confidence - inside and outside the organisation.
Organisations leading in this space set a clear leadership tone.
Leadership pressure will intensify. Stakeholder scrutiny will grow. Expectations for explainability will rise. And waiting for perfect clarity will become a risk in itself.
The organisations that thrive will be those whose leaders step deliberately into uncertainty and guide their teams through it.
Direction, clarity, and mindset matter more than maturity models or perfect plans.
Acting now makes organisations not just more productive, but more resilient - and positions leadership where it belongs: at the centre of transformation.
KPMG’s AI consulting team can help you navigate this transition.
