AI has moved from experiment to boardroom priority in a remarkably short time, yet far fewer organisations can point to AI that has genuinely changed how they compete, operate or create value. Closing that gap is fast becoming a defining test of leadership.
Findings from the Kestria Global Leadership Barometer 2026, based on responses from more than 700 leaders worldwide, show that around two thirds of organisations have either a developing AI strategy or a reactive one, while fewer than 15% describe their AI strategy as fully integrated.
The barriers they cite (insufficient internal capability, difficulty translating strategy into execution, cultural resistance and competing priorities) are leadership problems rather than technology problems.
Kestria interviewed senior executives from around the globe to explore how organisations are turning AI strategy into measurable business impact.
What distinguishes companies that are successfully turning AI ambition into business impact?
For Phil Eames, former Global Chief Operating Officer at Kantar Media, based in UK, successful organisations ask a different question. ‘Most companies are running pilots or rolling out productivity tools with little direction. The real question is where AI can materially change the economics, performance or competitive position of the business. AI must be part of business strategy, owned by the executive team and board, with essential technical leadership from the CTO and CIO. Rather than simply automating, ask whether you would design workflows, decisions, roles and structures the same way if AI had been available from the outset.
Transformation changes the operating model and delivers measurable outcomes in revenue, service levels or delivery costs. Leaders must test, learn and accept failures while maintaining focus on value. Scaling requires sound infrastructure, architecture, data, governance and security, supported by a platform to control and measure what is built.’
Evan Situ, Head of Technology for Data and AI at APPLY, based in Canada, highlights proprietary business context as a source of competitive advantage. ‘As AI models become increasingly accessible, differentiation depends on how effectively organisations combine them with proprietary data, domain expertise, institutional knowledge and unique business processes. Successful companies focus investment on a concentrated portfolio of opportunities tied to measurable business outcomes, using AI to rethink workflows rather than simply automate individual tasks.’
Giovanni Romero, Operating Partner at Providence Equity Partners, based in UK, points to rapid technological change and organisational design. ‘AI decisions made three or six months ago can already be obsolete, yet workflows still follow reporting lines. The greatest value comes from redesigning around workflows, defining where agents can act and humans must approve, with controls over prompts, agent operation and access to tools and data built into the platform. Remain model-agnostic, selecting AI for each use case according to capability, cost and platform.
Measurable results remain limited. We track hours saved rather than headcount, alongside direct EBITDA impact. Private equity isn’t waiting for perfect metrics: there is broad conviction that portfolio companies must become AI-enabled or AI-native. Our medium-term horizon makes that easier than in public markets, where daily, quarterly and annual performance pressures can conflict.’
Ganendran Pathmanathan, Head of Procurement Development, Asia Pacific at ABB Energy Industries Division, based in Malaysia, emphasises middle management’s role in translating AI strategy into operational impact. ‘While C-suite sponsorship is crucial, the real leverage lies in middle management—the people who redesign processes, manage operational execution and make daily decisions. Sustainable AI deployment depends on their ability to connect the organisation’s most pressing challenges with practical changes to its operating model, supported by disciplined execution and cross-functional alignment.’
Rob Straker, Executive Director, AI, Data & Analytics at KPMG Canada, based in Canada, sees adoption as highly uneven. ‘Many organisations are experimenting with AI embedded in existing tools rather than investing deliberately. Despite Canada’s pioneering role in AI, many businesses remain cautious. Those embedding AI in their strategy put innovation at their core and aspire to be world class. Shopify, for example, requires teams seeking additional headcount or resources to show why the work cannot be done with AI.
AI is as fundamental as IT was in the 1990s. Like electricity, its early productivity gains precede deeper changes to how businesses operate and the creation of entirely new businesses. That is why I encourage every organisation to start experimenting now.’
For Yann Dietrich, Group Head of AI, Data & IP – Legal at Atos, based in France, the companies that succeed distinguish between agentic AI and transformation. 'Agentic AI acts as an assistant: one I allow to make decisions for me, which I remain in control of, and with which I can deliver real impact. On its own it is useful, but it does not deliver productivity gains at scale. Real transformation comes from embedding AI into processes and transforming those processes around it.'
What role should executive leadership play in ensuring AI becomes a strategic advantage rather than simply another technology initiative?
‘AI is a team game, not an IT game,’ says Phil Eames. ‘Boards and executive teams generally have the capability but often lack confidence, delegating AI to technology leaders. They must remain involved in strategy, operating-model decisions, priorities, risk, governance and people. If you don’t self-disrupt, someone else will.
Establish governance early to prevent teams independently building agents with access to operational systems and proprietary data. Monitor quality and performance closely, increasing AI’s control as confidence grows. Job concerns are legitimate: routine roles may decline while demand for data and governance specialists increases, and work shifts towards testing, monitoring and improvement. Clear communication and urgency are essential, because fear creates resistance.’
Giovanni Romero sees AI as a shift comparable to the Industrial Revolution, electricity and computing. ‘Boards and executives need to understand its opportunities, risks and barriers, without mastering the technical mechanics. The CEO must lead, whether AI presents an existential challenge or an opportunity.
We support CEOs with training and specialist advisers, assess AI throughout the investment cycle, and expect every portfolio company to have a policy, plan, deployed tools and identified use cases. People leaders must also be involved: For example, deploying a headquarters AI tool in Spain and France required union consultation. The aim is not workforce replacement, but a business growing threefold should not need three times the headcount.’
Evan Situ draws a clear distinction between strategic ambition, business accountability and technical delivery. ‘The CEO must establish the ambition, business leaders must own the outcomes, and technology teams must enable secure, scalable and reliable execution. Executives must create the conditions for success through focused investment, AI fluency, redesigned roles and responsible governance. Technology alone cannot deliver transformation.’
For Rob Straker, changing how work gets done requires both top-down and bottom-up effort. ‘Leaders may be too far from daily operations to identify the best opportunities, while employee-led initiatives struggle without leadership support and funding. AI transformation is often a multi-year journey. Leaders must make AI a strategic priority, invest in education and initial tools, establish accountability, and develop a roadmap. In parallel, teams should identify high-value use cases, experiment, progressively scale what works and share lessons as capabilities mature.
Governance of emerging technologies like AI requires caution and humility: you cannot effectively govern what you do not yet understand. Allow experimentation with safeguards from the outset, adapting governance as understanding and deployment grow. Like the brakes on a Formula 1 car, good governance enables speed: move low-risk uses through quickly and scrutinise higher-risk ones more closely.’
Ganendran Pathmanathan emphasises the need for leaders to align teams and create incentives for successful AI deployment. ‘A key role of leadership is to break down silos, encouraging cross-functional collaboration and shared accountability. Building a reward structure around successful AI deployment can help build momentum and extend its impact across the wider organisation.’
Yann Dietrich believes leadership's first task is to help the organisation embrace AI. 'Removing the fear of being replaced by AI is a message that must come from the top of the organisation. People need to get comfortable with AI, and leaders need to calculate and demonstrate its value. AI should be used to advance capability and accelerate the business; it is not just about cost savings.'
What leadership qualities will define successful AI transformation over the next five years?
Phil Eames highlights pragmatism, the ability to operate within uncertainty, curiosity and courage. ‘Curiosity is often drummed out of organisations, and brave decisions won’t always work. I see three stages: AI as a productivity tool, embedded in end-to-end workflows, and integrated into the operating model. Most organisations remain at stage one; the real competitive advantage lies at stage three, redesigning products, services and processes around AI. Leaders must ask what their company should look like.
When deciding whether to build or buy, invest in processes that genuinely differentiate the business, giving capable individuals accountability for priority areas. For routine tasks such as CV sifting, someone has probably already solved the problem.’
For Giovanni Romero, the defining qualities are continuous development and curiosity. ‘Managing 500 developers for 20 years doesn’t guarantee success when your technology needs refactoring within two years. Leaders must continually adapt: leading a crew from a sailing boat to an electric yacht still requires people leadership, but different expertise. People leadership and clear goals remain essential for CEOs, while functional leaders face even greater changes.
There is no universal build-versus-buy playbook. Keep your competitive moat—knowledge, data and relationships—under your control. I’ve seen vendors use a company’s data to improve systems they then sell to its competitors. Standard functions can go to vendors, but rigorous due diligence is essential: some so-called agents are simply chatbot interfaces.’
Rob Straker believes successful AI leadership requires commercial judgement, adaptability and the ability to lead organisational change. ‘Returns extend beyond cost and efficiency to quality, speed, innovation, employee and customer satisfaction, and attracting talent. Establish a baseline, measure actual outcomes and redirect investment when benefits fall short. Fund foundational infrastructure as a shared strategic investment, rather than burdening individual use cases with its full cost.
AI is a general-purpose capability that can fundamentally change business models; treating it simply as an IT tool narrows thinking and pushes ownership towards the CIO. In one organisation I work with, AI is led by the marketing and product leader. Change management must also become a stronger capability, with HR involved early to keep transformation human-centred from identifying opportunities through deployment. Employees understand how the business works.’
Ganendran Pathmanathan highlights intellectual humility and ethical responsibility as essential leadership qualities. ‘Leaders will need a strong understanding of AI’s strengths, limitations and capabilities in a continuously changing landscape. They must also have the resolve to govern AI responsibly, even under pressure to move quickly and remain competitive.’
For Yann Dietrich, governance is fundamental. 'Leaders must operate within a level of risk that is acceptable to the organisation. The successful leaders in the age of AI will be orchestrators, because everyone will become a manager of AI.'
Evan Situ sees continuous reinvention as a defining leadership capability. ‘Leaders must connect technological possibilities with business economics and mobilise people through ongoing change. AI transformation is not a one-time programme: it requires organisations that continuously learn, experiment and evolve. The central question will shift from “How can we use AI in our business?” to “What should our business become now that AI makes entirely new ways of operating possible?”’
Conclusion
The Kestria Global Leadership Barometer 2026 shows that awareness of AI is widespread but maturity remains uneven, with fewer than 15% of organisations reporting a fully integrated AI strategy. The executives interviewed agree that what separates experimentation from transformation is leadership, not technology. Organisations that turn AI ambition into business impact anchor it in business strategy, pair top-down commitment with bottom-up innovation, redesign their operating models around it and measure it against a broad view of value. Ownership sits with business leadership rather than IT, supported by governance that enables pace and a workforce brought along with clarity rather than fear. The leaders who succeed over the next five years will combine curiosity, courage and continuous learning with the judgement to disrupt their own businesses before others do.
