AI Is Becoming a Teammate. Workforce Planning Has to Change.
New research on the cybernetic teammate suggests AI is not just changing productivity. It is changing team design, expertise, and workforce planning.
Last week, I joined a Wharton alumni discussion with Professor Gad Allon on one of the most important workforce questions facing executives:
What happens when AI stops behaving like a tool and starts behaving like a teammate?
The conversation centered on emerging research, including the field experiment The Cybernetic Teammate, conducted by researchers from Harvard, Wharton, and Procter & Gamble. The study looked at 776 P&G professionals working on real product innovation challenges, comparing individuals and teams, with and without AI.
The findings should have the attention of CHROs, CEOs, operating partners, and functional leaders.
The first signal was leverage.
Individuals using AI produced work comparable in quality to two-person teams working without AI. That does not mean the simplistic answer is “fewer people.” It means leaders have to rethink how work gets designed.
For decades, organizations assumed complex work required larger, more specialized teams. Coordination costs were treated as the price of solving hard problems. But if one strong employee paired with AI can generate work that looks more like team output, workforce planning cannot stop at headcount, spans, and static org charts.
Leaders will need to ask different questions:
- What work truly requires a team?
- What work can be handled by an AI-enabled individual?
- Which coordination layers still create value?
- Where should decision rights expand because capability has expanded?
The second signal was integration.
Organizations rely on teams partly because expertise is fragmented. Commercial leaders think commercially. Technical leaders think technically. Legal, finance, HR, operations, and product each bring a different lens.
Without AI, the research showed predictable functional bias. With AI, those boundaries became less rigid. Individuals using AI generated more balanced, cross-functional solutions on their own.
That matters because AI may become a boundary-spanning capability. It can help employees reason beyond their native domain, test assumptions, translate ideas across functions, and access expertise in real time.
The implication is not the elimination of expertise. It is the democratization of expertise.
That should change how organizations think about capability building. The future skill is not simply prompting. Prompting is table stakes. The real skill is judgment: framing better questions, validating outputs, integrating multiple perspectives, applying context, and knowing when human expertise still needs to own the decision.
The third signal was emotional.
The dominant AI narrative often centers on anxiety: displacement, burnout, isolation, and fear. But participants using AI reported higher enthusiasm, energy, and excitement, and lower anxiety and frustration.
That is important because adoption is rarely just a technology problem. It is a behavior change problem.
Employees are more likely to adopt AI when it helps them feel more effective, more confident, and more capable at work. Organizations that frame AI only as cost reduction may create resistance. Organizations that frame it as a capability amplifier are more likely to create durable adoption.
The workforce planning model has to evolve.
The old question was: How many people do we need?
The better question is: What combination of humans, AI, expertise, judgment, and governance is required to achieve the outcome?
That shift changes team design. It changes role design. It changes management. It changes how leaders think about productivity, quality, risk, and innovation.
One nuance from the research is especially important: AI-enabled individuals may create efficiency, but teams using AI were more likely to generate exceptional, top-decile ideas.
That distinction is the future of work in one sentence.
Efficiency may come from AI-enabled individuals. Breakthrough innovation may come from human collaboration amplified by AI.
The question for executives is no longer whether AI changes work. It already has.
The real question is whether we are redesigning the workforce fast enough to take advantage of it.