As AI agents take on more autonomous work, organisations must rethink leadership, redesign jobs and invest in workforce skills to ensure people remain at the heart of successful transformation.
AI agents will reshape how the world works, changing styles of leadership along with it, with people, skills and trust at the centre.
Agents are autonomous, proactive, digital employees that can reason, make decisions and take independent actions to support people and complete complex business processes.
The organisations utilising these programs share a common trait: even at this early stage of their agentic transformation, they are proactively investing in workforce redesign. By treating AI fluency as an organisation-wide capability rather than a specialist skillset, they are redefining how work can be done.
At Salesforce, we learned this firsthand when we deployed Agentforce. It handled 2.6 million unanswered customer queries. More importantly, the hundreds of support engineers redeployed into new roles became, according to our hiring managers, some of the highest performers at the company.
Today, 100% of Salesforce employees are engaging with AI in some way, reinforcing a core lesson: transformation works when people are given the confidence, context and tools to use AI in their day-to-day work. And it’s not just us. Our survey of global HR executives found that Chief Human Resource Officers expect to redeploy nearly a quarter of their workforce as AI agent adoption grows.
That experience has shaped our AI Fluency Playbook, a practical framework anchored in the 4Rs – Redesign, Reskill, Redeploy and Rebalance. This helps leaders redesign work, build AI fluency across their workforce and ensure people remain central to the agentic enterprise.
Two principles for workplace transformation
Before a company can redesign its workforce, leaders have to be clear about what they are actually designing for. That clarity begins with two foundational principles – the ability to speak confidently and practically about why AI fluency matters at every level of the organisation.
To begin with, humans remain in control. AI agents are powerful because they can act autonomously at scale, but that autonomy must be governed, auditable, and aligned with human intent. Leaders who rush past this principle aren’t just risking errors, but eroding the trust of their employees and customers.
Additionally, the goal is augmentation, not replacement. Agents handle work that benefits from speed and scale, like processing high volumes of incoming service requests, surfacing the right information, and following up on routine queries. Humans focus on work that demands judgment, creativity and contextual understanding. These include navigating a difficult customer conversation or building the kind of trust that no automated interaction can. Neither replaces the other. Together, they accomplish what neither could alone.
These factors only become operational when employees are supported to build the confidence and judgement to work effectively with AI.
A framework for action: the 4Rs
With those principles as the foundation, leaders can move to the harder part: redesigning how work is completed. The most effective transformations follow a clear sequence, which we describe through the 4Rs.
Redesign involves examining work at the task level and moving from scripted processes to outcome-based workflows. Rather than layering AI on top of existing roles, define the desired end goal first and identify where human judgment is non-negotiable. This requires leaders to question how work has always been done.
Reskill is another crucial element. In Ireland, the PwC AI Jobs Barometer found that demand for AI-skilled jobs has grown by 83% since 2019, more than five times faster than the overall jobs market. That is not a future problem. Build AI fluency, data literacy, critical and systems thinking into workforce development programmes now, during working hours, not as an optional extra. The goal is not just to train people on tools, but to help them adapt, question outputs and apply human judgement with confidence.
Redeploy requires auditing your existing workforce for transferable capability before posting a new role. The ability to navigate complexity, stay composed under pressure, and solve problems with incomplete information is precisely what high-value, AI-augmented positions require.
For rebalancing, you must establish a clear and ongoing division of responsibility between human and agent, and revisit it regularly as both capabilities evolve.
This same challenge extends beyond today’s workforce. If we want the next generation to thrive in an AI-shaped economy, education systems must place greater emphasis on the durable skills that AI cannot replace: critical thinking, adaptability, communication and the ability to learn continuously. Students will need to work with AI tools that are still evolving, which means preparing them for change is just as important as preparing them for any single technology.
The leadership imperative
AI will not transform your company – your people will, but only if you give them the opportunity.
The transition to an agentic way of working is not frictionless. It asks employees to learn new skills, build confidence in unfamiliar territory and trust that their existing expertise has a place in the future. What makes the difference is leadership that invests in the conditions that allow people to grow: dedicated time for training, active manager support and a clear sense of direction.
The cost of retaining your people, their knowledge, experience and institutional expertise, is almost always less than the cost of losing them. Leaders who recognise this early will have a durable competitive advantage: a workforce that is more agile, more capable and more engaged because they were supported through the transition, not sidelined by it.
Invest in your people’s next chapter as heavily as you invest in the technology itself. That is the surest path to an AI transformation that works.
