Beyond Headcount: Designing the Human-AI Workforce

Artificial intelligence is rapidly changing how individuals work. But for organizations, the bigger question may be how AI changes work itself.

As AI and AI agents become embedded in workflows, leaders have new decisions to make: What work is best performed by people? What can AI take on? Where should the two work together? And how does that change the way organizations structure teams, plan their workforce and develop talent? Few leaders are better positioned to help answer those questions than talent development and HR leaders.

I recently explored the implications with Cathy Manginelli, who brings nearly three decades of experience in talent management, including roles as Chief Talent Officer at TriNet, Managing Director of Talent at TD Ameritrade and in global talent leadership across the financial services industry.

Our conversation centered on a fundamental shift: moving beyond using AI to make today’s workforce more productive toward intentionally designing organizations for a workforce that increasingly includes both humans and AI.

LISA: There’s much discussion about how AI will change individual jobs. Are we thinking big enough? How do you see AI changing the design of organizations themselves?

CATHY: Individual jobs changing as a result of new technology is not a new concept. Technology has always been evolving in our workflows, whether moving from manual tasks to self-service or introducing chatbots for first-level communications and service. Technology evolution has always forced organizations to rethink the way work gets done.

I think the difference now is the pace and speed. AI and generative AI are driving us to transform at a pace most organizations are not prepared to absorb or proactively design for.

AI will change the shape of organizations. The real shift is architectural. AI is pushing organizations toward new operating models, new structures and new ways of coordinating human effort. Before AI agents, organizational models followed a pattern of strategy, structure and staffing. Now, structure and staffing are intertwined and represent not only humans but technology — agents — as part of the workflow and workforce design.

That introduces new questions about how to lead and manage an agent versus a human, how to direct the work, and how humans and agents collaborate and partner. It could ultimately flatten the hierarchy. With agents in the organizational design, work can get done faster, decisions can be made faster, predictive modeling and analytics become more readily available, and information and data can flow more freely.

However, what I have found from my own experience using agents is that you still need the human in the loop. Agents, at this stage, still need support with critical thinking. They have little emotional intelligence. They can’t “read a room” or pick up on cues to understand what is not being said but should be considered for higher-order work.

The implication is that, just like humans, you need the right agent in the right role at the right time. You have to be very clear about the work that agent needs to do, how it will collaborate, and how it will get and give information, and then program it accordingly. Agents are not perfect or foolproof and, like humans, need oversight, direction, support, reinforcement and coaching.

AI can also help us design better organizational structures. It can more quickly model different spans, layers and structures; assess communication flows and barriers to workflow; and help better align structure, skills and strategy. Predictive models can highlight gaps that organizational charts alone won’t show and reveal talent gaps, scarcity and critical-role risk.

LISA: As AI becomes capable of taking on more tasks, how should CHROs rethink workforce planning — not just how many people they need, but what work needs to be done and how it should be organized?

CATHY: CHROs need to become very conversant in the responsibilities required, not just the role. This may require doing time and task studies, observing, testing, piloting and decomposing the actual job. They will need to assess the decision rights required and determine what a human needs to deliver versus what AI can execute. Tasks requiring critical thinking, EQ, empathy, contextual judgment, creativity and originality are best suited for a human.

We must abandon the notion of a traditional job description and move toward a job profile with clear outcomes and AI collaboration requirements.

The term “hybrid workforce” typically refers to where a worker is working. Now it also means who (or what) is doing the work. Is it an AI agent or a human? CHROs need to define where AI acts as:

• Autonomous performer, such as payroll anomaly detection
• Co-pilot, such as manager performance-review drafting
• Advisor, such as predictive people analytics
• Orchestrator, such as routing HR service-desk queries

It’s no longer about headcount planning but capability planning. What technical, human and system skills are needed? From there, organizations can build the talent plan and workforce structure for agility, allowing networked teams and cohorts to come together when needed to deliver.

LISA: People have traditionally represented one of an organization’s largest operating expenses, and companies are now making significant investments in AI as well. How should leaders think about those investments together rather than as separate technology and talent decisions?

CATHY: Leaders should look at the entire ecosystem of productivity in a way that represents both operating expense and capital expense. Both are incredibly important to the long-term health and operating success of the organization, but leaders need to understand how one is delivering immediate value and the other is a more enduring utility.

The frame I think of is the total workforce. We have employees and contractors/contingent workers. We can look at them separately, but if we do, the entire picture is not clear. We have to consider the total workforce as we look at our strategies and investments.

I see AI the same way. It has to be part of the financial equation. “Cost centers” need to be regarded as capability centers. The classification comes from traditional accounting rules, which will need to change and evolve if we are to combine the two concepts.

LISA: Where do you think the bigger opportunity lies: using AI to make today’s workforce more productive, or redesigning work around what humans and AI can each do best?

CATHY: I think both are equally important strategic choices. However, if you try to plug AI into an old organizational model, it won’t work. You have to start by redesigning the work first. Then you determine: Is that work best for a human or AI?

The redesign is about building an organizational architecture that is flexible, scalable and responsive. It can’t be rigid. It has to allow for quick learning and evolution. Leadership needs to understand how to be fluid now more than ever.

LISA: If AI is going to fundamentally change how work gets done, what role should the CHRO play, and what should CHROs be doing now rather than waiting for the technology to evolve further?

CATHY: The CHRO needs to become the chief work architect or designer. Their role is all about talent and workforce design and development. CHROs and other talent leaders should build a human-AI capability model, create a governance framework for AI-enabled work, and communicate a clear narrative about the human in the loop and what the future of work looks like.