Human-powered reinvention: a deep dive

Q1: Why is now a key moment for businesses to determine their approach to Agentic AI? 

We are at an inflection point because Agentic AI is no longer limited to assisting individual tasks. Agentic AI can plan, make decisions, and execute across end-to-end processes. That fundamentally changes how work gets done. 

If organisations treat this as just another automation wave, they risk locking themselves into operating models that assume humans remain the primary coordinators of work. As Agentic AI matures, those assumptions become constraints. Delaying a clear strategy does not preserve flexibility. It reduces it. Leaders need to decide now how much of their future organisation will be designed around human-only workflows, and how much will be rebuilt for human–agent collaboration.

Q2: What are the risks of focusing on automation through AI rather than reinvention? 

The biggest risk is optimising a model of work that is already becoming obsolete. Automating existing processes can deliver short-term efficiency gains, but it also reinforces legacy assumptions about roles, decision rights, and control structures. 

Agentic AI challenges those assumptions. It can coordinate work, adapt in real time, and operate across silos of humans and data in ways traditional hierarchies were never designed to support. Organisations that only automate end up layering AI on top of processes that no longer make sense, rather than redesigning how value is created. Reinvention requires leaders to rethink workflows, roles, and governance from the ground up, not simply make today’s model run faster. 

Q3: The article talks about leaders needing ‘the courage to make decisions that may feel uncomfortable in the short term, in order to create long-term opportunity.’ Can you share some insight as to how leaders can achieve this balance? 

We are seeing some organisations adopt a dual-track approach. One track focuses on running today’s business more efficiently, using AI to incrementally improve performance. The other operates almost like an internal startup, experimenting aggressively with new AI-enabled ways of working and new business models. 

This second track often challenges existing power structures, roles, and revenue streams, which is uncomfortable by design. But it allows leaders to explore what comes next before the current model starts to break down. The balance comes from accepting that part of the organisation is investing in its own disruption, rather than pretending transformation can be achieved without trade-offs. 

Q4: You’ve mentioned that Agentic AI is proactive – with this enhanced capability, what will the ‘new models of human-agent collaboration’ look like? 

Human–agent collaboration will shift the centre of gravity of work. People will increasingly focus on judgement, context, ethics, and problem framing, while agents handle coordination, data synthesis, and large-scale execution. 

This changes how teams operate day to day. Decisions can move faster, work can scale without linear increases in headcount, and collaboration becomes less about managing handoffs and more about guiding outcomes. As a result, we are already seeing the emergence of new roles focused on designing, supervising, and improving human–agent systems, not just performing individual tasks within them.

Q5: How can leaders build trust among employees around the potential disruption that Agentic AI brings with it? 

Trust is built through honesty and participation, not reassurance. Leaders need to involve employees early in shaping how AI is introduced, and be clear about what will change as well as what will not. Treating people as passive recipients of technology decisions undermines credibility very quickly. 

Building trust also means backing words with action. Investing in reskilling, creating transparent governance around how AI is used, and being explicit about decision principles all signal that AI is being deployed deliberately, not opportunistically. Employees are more likely to engage when they understand the direction of travel and their role within it, even when the change is disruptive. 

Q6: You’ve mentioned new roles being created – such as Agentic Process Designers – but their lagging behind the development and adoption of AI itself. How can businesses catch up? 

Organisations need to stop treating roles as fixed job descriptions and start treating them as evolving design choices. Agentic AI changes workflows faster than traditional workforce planning cycles can keep up with. 

One effective approach is to use pilot programmes to reinvent processes with AI and observe how roles actually shift in practice. That evidence can then inform new role definitions, skill requirements, and career paths. This needs to be supported by robust data on existing skills and future needs, as well as partnerships with academic and industry bodies. The goal is not to predict the perfect role upfront, but to learn and adapt faster than the technology itself is changing. 

Q7: Agentic AI challenges leaders to shift from aiming to resolve tension to learning to live inside it. Evidently, future leaders will need different skills to the past…but will there be a need for a different type of leader? 

Yes. Leadership will move away from directing tasks and toward enabling adaptation. In an environment where AI systems can act autonomously, leaders cannot rely on control and certainty as their primary tools. 

Future leaders will need to be comfortable operating amid ambiguity, managing competing priorities, and learning through rapid experimentation. Rather than resolving tension quickly, they will need to hold it, recognising that uncertainty is not a failure state but a feature of transformation. Their role becomes less about having the answers and more about creating the conditions for the organisation to learn and evolve continuously. 

Q8: Finally, if a leader could only take away one key learning from the article, what would you like that to be? 

Do not mistake automation for transformation. Agentic AI’s real value lies not in making existing work cheaper or faster, but in enabling organisations to rethink how work is structured, how roles are defined, and how leadership is exercised. 

Leaders who focus only on efficiency will capture incremental gains. Those who are willing to redesign their organisations around human–agent collaboration can unlock entirely new sources of value. That choice needs to be made deliberately, and sooner rather than later.

This webpage includes AI generated content