by Lisa Dean, EMA’s Head of Transformation & Marketing
The defining leadership challenge of artificial intelligence is not simply deciding which tools to adopt. It is guiding an organisation through uncertainty while technology, customer expectations and work itself are changing.
Leaders are being asked to make decisions about workforce transformation, governance, risk and organisational design before the implications are fully understood. They must provide direction without pretending to have every answer, balancing the need to move with the responsibility to move wisely.
This is the leadership gap emerging through Workforce 2030, the EMA’s research and capability-building initiative focused on the future of work. Technological capability is advancing faster than the organisational and leadership capability needed to apply it well.
The EMA’s Workforce 2030 research found that 83% of surveyed businesses were using AI in some form. One in four had embedded it into the way they operated, while many others were testing tools and working out what AI meant for their people.
The next challenge is turning that activity into better business performance. That requires more than technology. It requires leaders who can make sound decisions in an evolving environment, build confidence across their organisations and redesign work around the value they want to create.
Leading when the answers are still emerging
Leadership confidence does not mean certainty. It means acting with sound judgement when the evidence is incomplete, learning from what happens and adjusting course. This is adaptive leadership in practice: providing direction without claiming certainty, learning through experimentation, adjusting as conditions change and helping people navigate that change.
That is a different way of leading from the traditional model in which seniority is associated with having the answers. AI can rapidly produce information, identify patterns and generate possible responses. Access to information is no longer the scarce resource it once was.
What matters is how leaders interpret that information. They must decide what to trust, what questions have not been asked and where human oversight is required. They must balance commercial opportunity with risk, ethics and organisational values.
They also need to create space for experimentation. Rather than betting everything on a single large-scale solution, organisations can test defined applications, establish clear guardrails, assess the results and apply what they learn.
Governance enables this progress. Good governance gives people the boundaries, accountability and confidence to experiment responsibly. It helps an organisation move forward without losing sight of the customers, employees and communities affected by
AI can inform decisions, but leaders remain accountable for the choices they make.
Redesign the work, not just the process
Many organisations are using AI to digitise existing work. They are introducing faster tools while leaving the underlying processes, roles and decision rights largely untouched.
That may produce incremental gains, but it misses the larger opportunity. The more useful question is: if we were designing this work today, with the technology now available, how would we structure it?
Work redesign begins with the outcome, not the tool. What is the organisation trying to achieve for its customers? Which capabilities are needed to deliver that outcome? Where can technology remove low-value effort or improve the quality and speed of a service? Where are human expertise, relationships and accountability essential?
This shifts the leadership conversation from automating tasks to organising capability. That capability may include employees, contractors, specialists, digital tools, automated systems and AI agents. Work may no longer fit neatly within established job descriptions or organisational boundaries.
Customer expectations must remain central. Faster internal processes have limited value if they do not improve the experience, responsiveness or value an organisation provides. The purpose of redesign is not to use more technology. It is to build a better-performing organisation around the needs it exists to serve.
The people doing the work should be closely involved. They often know where duplication, frustration, unnecessary hand-offs and wasted effort sit. Their practical knowledge is essential to designing something better and to identifying consequences that may not be visible at the executive level.
Leadership responsibility is shared
No single executive can own this change.
CEOs and CPOs share responsibility for ensuring AI delivers organisational value. Together they must align strategy, investment, workforce capability, work redesign and culture. Technology leaders must provide reliable systems and explain their limitations. Operational managers must translate organisational intent into everyday decisions and working practices.
Middle managers have a particularly important role. They sit closest to where change lands and are often the bridge between organisational ambition and day-to-day reality. Employees need more than an explanation of why AI is being adopted. They need clarity on what it means for their work, their customers and the decisions they make.
Organisations cannot expect managers to lead this change effectively without giving them the information, authority and development they need.
The strongest leadership teams will treat AI adoption as a shared organisational responsibility, with clear decision rights and genuine collaboration across strategy, people, technology and operations.
The commercial test
The ultimate measure of any investment in change is not how many new tools an organisation adopts. It is whether those investments improve productivity, innovation, customer value, competitiveness and sustainable growth.
That outcome will depend heavily on leadership capability.
The organisations that realise the greatest value will be those that can learn, adapt and evolve while maintaining clear strategic direction. They will create safe opportunities to experiment, involve employees in redesigning work and build the confidence needed to turn uncertainty into progress.
Their leaders will not be expected to have all the answers. They will be expected to create clarity when circumstances are changing and help people navigate ambiguity. They will also need to build organisations that can adapt as expectations, roles and ways of working continue to evolve.
The future will not belong to organisations that can predict every change. It will belong to those that can respond effectively when change arrives.
That remains deeply human work.