From Headcount to Human & AI Capacity

Strategic Workforce Planning in the Age of AI Agents

Many companies face the same challenge: Traditional workforce planning no longer works once AI not only provides support but also takes on tasks independently. Let’s therefore imagine a company that views AI agents not as a tool, but as a plannable capacity.

In addition to traditional FTEs, so-called “agent slots” emerge. An agent slot describes a plannable capacity unit for an AI agent, with specific tasks, permissions, and performance metrics.

Role profiles and skill architectures are harmonized across the organization: from human-in-the-loop reviewers to roles for managing AI agents, all the way to responsibilities for security, compliance, and governance.

As a result, strategic workforce planning shifts from the question “How many people, in which roles?” to “What mix of people and AI delivers the right quality, speed, and resilience that is both regulatory-compliant and culturally sustainable?”

Why traditional strategic workforce planning is reaching its limits

Traditional workforce planning manages capacity through roles, FTEs, and headcounts. This model assumes stable processes in which value creation is almost exclusively performed by humans.

However, AI accelerates processes, takes over sets of tasks, and shifts responsibilities. Rigid staffing plans react too slowly to these dynamics. At the same time, it remains unclear in many areas which tasks can be reliably and fully automated and where human responsibility remains absolutely essential.

Studies by McKinsey and the World Economic Forum show: Organizations rarely fail with AI because of the technology itself, but rather due to unclear roles and a lack of skills. Workforce planning of the future will therefore no longer just manage people, but will shape the distribution of responsibility between humans and machines. Ideally, this will be done in close collaboration with IT. This is because tasks, roles, and ways of working are changing particularly rapidly in this area due to AI, and the understanding of technology, automation, and their limitations is greatest here.

How work is redistributed and AI agents become part of the team

To plan Human & AI Capacity, we must first examine the work itself. On this basis, work can be divided into three categories:

• Empathy and responsibility: Tasks that are deliberately kept human because they require trust, context, responsibility, or personal presence.
• Human-AI collaboration: AI agents take on preparatory, structuring, or analytical tasks, while humans provide context, steer the process, and take responsibility for decisions.
• Full AI automation: Highly standardized, data-driven processes that can be executed autonomously by AI according to clear rules.

Depending on technological maturity, regulatory frameworks, and quality standards, tasks shift between these categories. Agent slots make these shifts visible and manageable: From a planning perspective, an agent slot can be treated similarly to a part-time position, except that it is flexibly scalable, technically interchangeable, and available 24/7. This allows teams to consciously decide which task bundles should be permanently automated, which should be anchored in human-AI collaboration, and which should remain explicitly “human only.”

As a result, teams are also changing: AI agents take on clearly defined tasks and, as agent slots, become functional team members that reduce human workload. In capacity planning, teams thus no longer think solely in terms of headcount but also in terms of the configuration of their agent slots, for example for pre-qualification, document classification, or internal assistance. In everyday life, these agents become small personal assistants. For instance, during a job interview, one might add with a smile: “I have my agent team with me; they’re currently working on a new social media campaign.”

The stronger this interaction becomes, the more important transparency, clear accountability, and robust control mechanisms become. Each agent slot requires defined responsible parties, key performance indicators, and escalation procedures. That is why companies are establishing organizational guidelines to manage the use of AI and agent slots, ensure quality, and regulate escalations. The insurance companies Allianz and ERGO, for example, have established central AI governance structures for this purpose that consolidate deployment, quality, and escalations.

How the target vision takes shape in practice

What is described as the target vision can already be observed today—albeit with varying degrees of maturity.

The Chinese insurer ZhongAn achieves an exceptionally high ratio of policies to employees, often in the six-figure range per person. This scale is achieved through radical automation. Instead of handling individual cases, the team manages AI-supported processes in underwriting, policy issuance, and claims settlement. This model illustrates the technical potential but requires a radically different understanding of operations and roles.

European insurers are taking a significantly more evolutionary approach. With its AI assistant Emmie, R+V Versicherung is specifically focusing on human-AI collaboration in customer service. Standard inquiries are answered automatically, while complex or sensitive issues are deliberately handed over to employees.

Bayerische is taking a similarly pragmatic approach. It is gradually transitioning AI from pilot projects into regular operations, for example in application support, email processing, or as an internal assistant for knowledge work. Employees gain time for consultation, analysis, and quality assurance, while AI takes on preparatory and structuring tasks. Organization and processes are deliberately co-developed in this process, rather than adapted retrospectively.

The connection between human and AI capacity, organizational design, and human resources development is particularly consistent at the ERGO Group. ERGO explicitly views AI as an organizational and people-related issue. Tasks are systematically redesigned so that AI takes on repetitive and preparatory tasks, while employees focus on complex, empathetic, and responsible activities.

In addition, ERGO is making targeted investments in expanded career paths. Through reskilling programs and its own academy, employees are prepared for roles in which they collaborate with AI, evaluate results, or take responsibility for its deployment. Qualification programs are prioritized based on data—that is, expanded where tasks are actually changing due to automation.

These examples show how insurers are not only automating work but also consciously shaping it. The result is adaptive operating models in which organization, work, and development are managed in an integrated manner.

What Human & AI Capacity Means for Organizations and Employees

For Human & AI Capacity to take root in the organization, new role models and clear responsibilities are needed. New roles are emerging for both people and AI. Strategic workforce planning must identify both early on, clearly define their roles, and make them visible within the organizational structure.

Examples of new human roles include:

• AI Development Manager (managing AI-centric initiatives)
• AI Integration Specialist (embedding AI into processes and systems)
• Human-in-the-loop roles (quality assurance and decision-making responsibility)

In parallel, AI agents assume specialized functional roles in the form of agent slots, each assigned to a responsible owner.

Skills development is also a key lever. Skill gaps must be identified and prioritized based on data. Not all employees require the same AI competencies.

What matters is who uses AI, who controls it, and who is responsible for it. This creates an operating model that can continuously adapt to technology, the market, and regulation. Work, organization, and development are intertwined. The focus is not on downsizing, but on targeted further development.

Many of the new roles are emerging around IT. That is why strategic workforce planning should start precisely where AI is developed, integrated, and controlled. We support insurers in building Human & AI Capacity in a structured way: We analyze work and roles, translate specific use cases into agent slots, identify AI and skill gaps, and derive sustainable organizational and training models from these findings. Our goal is to align FTEs, agent slots, and structure in such a way that change remains manageable and employees are provided with genuine development opportunities.