We will be the last generation to work with all-human workforces. This is not a provocative soundbite but a statement of fact, one that signals a fundamental shift in how organisations operate and what leadership now demands. The challenge facing today’s leaders is not simply adopting new technology but architecting an entirely new operating model where humans and autonomous AI agents work in concert.
According to Salesforce 2025 CEO research, 99% of CEOs say they are prepared to integrate digital labor into their business, yet only 51% feel fully prepared to do so. This gap between awareness and readiness reveals the central tension of this moment: we recognise the transformation ahead but lack established frameworks for navigating it. The question is no longer whether AI agents will reshape work, but whether leaders can develop the new capabilities required to direct this dual workforce effectively.
The scale of change is already visible in the data. According to the latest CIO trends, AI implementation has surged 282% year over year, jumping from 11% to 42% of organisations deploying AI at scale. Meanwhile, the IDC estimates that digital labour will generate a global economic impact of $13 trillion by 2030, with their research suggesting that agentic AI tools could enhance productivity by taking on the equivalent of almost 23% of a full-time employee’s weekly workload.
With the majority of CEOs acknowledging that digital labor will transform their company structure entirely, and that implementing agents is critical for competing in today’s economic climate, the reality is that transformation is not coming, it’s already here, and it requires a fundamental change to the way we approach leadership.
The Director of the Dual Workforce
Traditional management models, built on hierarchies of human workers executing tasks under supervision, were designed for a different era. What is needed now might be called the Director of the dual workforce, a leader whose mandate is not to execute every task but to architect and oversee effective collaboration between human teams and autonomous digital labor. This role is governed by five core principles that define how AI agents should be structured, deployed and optimised within organisations.
Structure forms the foundation. Just as organisational charts define human roles and reporting lines, leaders must design clear frameworks for AI agents, defining their scope, establishing mandates and setting boundaries for their operation. This is particularly challenging given that the average enterprise uses 897 applications, only 29% of which are connected. Leaders must create coherent structures within fragmented technology landscapes as a strong data foundation is the most critical factor for successful AI implementation. Without proper structure, agents risk operating in silos or creating new inefficiencies rather than resolving existing ones.
Oversight translates structure into accountability. Leaders must establish clear performance metrics and conduct regular reviews of their digital workforce, applying the same rigour they bring to managing human teams. This becomes essential as organisations scale beyond pilot projects and we’ve seen a significant increase in companies moving from pilot to production, indicating that the shift from experimentation to operational deployment is accelerating. It’s also clear that structured approaches to agent deployment can deliver return on investment substantially faster than do-it-yourself methods whilst reducing costs, but only when proper oversight mechanisms are in place.
To ensure agents learn from trusted data and behave as intended before deployment, training and testing is required. Leaders bear responsibility for curating the knowledge base agents access and rigorously testing their behaviour before release. This addresses a critical challenge: leaders believe their most valuable insights are trapped in roughly 19% of company data that remains siloed. The quality of training directly impacts performance and properly trained agents can achieve 75% higher accuracy than those deployed without rigorous preparation.
Additionally, strategy determines where and how to deploy agent resources for competitive advantage. This requires identifying high-value, repetitive or complex processes where AI augmentation drives meaningful impact. Early adoption patterns reveal clear trends: according to the Salesforce Agentic Enterprise Index tracking the first half of 2025, organisations saw a 119% increase in agents created, with top use cases spanning sales, service and internal business operations. The same research shows employees are engaging with AI agents 65% more frequently, and conversations are running 35% longer, suggesting that strategic deployment is finding genuine utility rather than novelty value.
The critical role of observability
The fifth principle, to observe and track, has emerged as perhaps the most critical enabler for scaling AI deployments safely. This requires real-time visibility into agent behaviour and performance, creating transparency that builds trust and enables rapid optimisation.
Given the surge in AI implementation, leaders need unified views of their AI operations to scale securely. Success hinges on seamless integration into core systems rather than isolated projects, and agentic AI demands new skills, with the top three in demand being leadership, storytelling and change management. The ability to observe and track agent performance is what makes this integration possible, allowing leaders to identify issues quickly, demonstrate accountability and make informed decisions about scaling.
The shift towards dual workforce management is already reshaping executive priorities and relationships. CIOs now partner more closely with CEOs than any other C-suite peer, reflecting their changing and central role in technology-driven strategy. Meanwhile, recent CHRO researchfound that 80% of Chief Human Resources Officers believe that within five years, most workforces will combine humans and AI agents, with expected productivity gains of 30% and labour cost reductions of 19%. The financial perspective has also clearly shifted dramatically, with CFOs moving away from cautious experimentation toward actively integrating AI agents into how they assess value, measure return on investment, and define broader business outcomes.
Leading the transition
The current generation of leaders are the crucial architects who must design and lead this transition. The role of director of the dual workforce is not aspirational but necessary, grounded in principles that govern effective agent deployment. Success requires moving beyond viewing AI as a technical initiative to understanding it as an organisational transformation that touches every aspect of operations, from workflow design to performance management to strategic planning.
This transformation also demands new capabilities from leaders themselves. The skills that defined effective management in all-human workforces remain important but are no longer sufficient. Leaders must develop fluency in understanding agent capabilities and limitations, learn to design workflows that optimally divide labor between humans and machines, and cultivate the ability to measure and optimise performance across both types of workers. They must also navigate the human dimensions of this transition, helping employees understand how their roles evolve, ensuring that the benefits of productivity gains are distributed fairly, and maintaining organisational cultures that value human judgement and creativity even as routine tasks migrate to digital labor.
The responsibility to direct what comes next, to architect systems where human creativity, judgement and relationship-building combine with the scalability, consistency and analytical power of AI agents, rests with today’s leaders. The organisations that thrive will be those whose directors embrace this mandate, developing the structures, oversight mechanisms, training protocols, strategic frameworks and observability systems that allow dual workforces to deliver on their considerable promise.
- Linda Saunders, Country Manager & Senior Director Solution Engineering Africa at Salesforce
