The New Human Capabilities L&D Teams Need in the Age of AI

As artificial intelligence begins to automate more of the drafting, generating, summarizing, and structuring work inside workplace learning, the capabilities that matter most in learning and development are starting to shift. What once distinguished many L&D professionals was their ability to produce high-quality content, structure learning experiences, and build instructional assets with speed and consistency. Increasingly, however, value is moving toward a different set of human strengths: framing, judgment, contextualization, orchestration, and oversight.

Drawing on patterns observed across seven large organizations adopting AI in workplace learning, this article explores how the capability profile of enterprise L&D is evolving and why legacy competency models are no longer sufficient for the work ahead.

This article draws on emerging findings from a joint research initiative by CommLab India and researchers at Lancaster University exploring how artificial intelligence is shaping workplace learning across large organizations. The patterns discussed here are informed by anonymized interviews with enterprise learning leaders across multiple industries and AI adoption contexts.

One of the most common assumptions about AI in workplace learning is that it is primarily a productivity tool. That assumption is not wrong, but it is incomplete.

Across enterprise learning teams, AI is already helping reduce the time required to draft content, generate assessments, summarize source material, create scenarios, adapt content across formats, and even support forms of role-play and practice. In practical terms, this means that tasks which once demanded substantial cognitive effort from learning professionals can now be completed more quickly, and in some cases, partially automated.

But beneath this visible efficiency layer, a deeper shift is underway.

AI is not simply changing how fast work gets done. It is changing what professional expertise in L&D actually looks like.

For years, many learning roles have been evaluated, both formally and informally, by the ability to produce learning outputs well. Could you write the storyboard, build the module, structure the workshop, shape the scenario, or turn SME input into something coherent and useful? These were not just deliverables. They were markers of expertise.

As AI begins to handle more of that first-pass production work, the basis of professional value begins to move.

This does not make human capability less important. It makes different human capabilities more important.

The question is no longer simply, “What can AI do for the learning team?”
The more strategic question is, “What must the learning team now become better at because AI is part of the system?”

That is where the real capability shift begins.

Across the seven organizations studied, AI was not replacing learning teams outright. But it was consistently absorbing pieces of cognitive work that had once required significant human effort and expertise.

This included:

On the surface, this may appear to be a simple efficiency story. But the deeper implication is that the relative value of different human contributions begins to change.

When AI handles more of the initial production work, people are increasingly needed not for volume, but for direction, interpretation, refinement, and responsibility. The work becomes less about creating everything from scratch and more about making better decisions inside a more automated system.

That is why the future of L&D capability is not simply about becoming “good at AI tools.” It is about becoming stronger in the distinctly human capabilities that make AI useful, safe, relevant, and strategically valuable.

How AI Is Shifting the Human Value Equation in L&D

Across the research, five capabilities emerged repeatedly as increasingly important in AI-enabled learning environments. These capabilities are not entirely new. But their relative importance is growing because AI is taking over more of the lower-friction production work that once absorbed so much of the learning team’s time and attention.

The ability to define the problem before the system generates the answer

AI is often celebrated for how quickly it can generate output. Yet in practice, the usefulness of that output depends heavily on how well the underlying problem has been framed. This is where human capability becomes indispensable.

Across the organizations studied, the teams that saw the strongest results from AI were not necessarily the teams using the largest number of tools. They were the teams that could define, with clarity and discipline:

This matters because enterprise learning problems are rarely just content problems. More often, they sit inside a much broader web of performance expectations, stakeholder needs, compliance realities, business constraints, and workflow conditions.

AI can generate quickly. It still depends on humans to define what problem is worth solving in the first place.

Without strong framing, AI tends to produce what looks like progress but is often just output. It may create more content, more options, or more polished drafts, but not necessarily the right intervention.

Why framing matters more now

In an AI-enabled environment, the ability to frame well is no longer a soft skill. It is a strategic capability.

The ability to decide what to trust, revise, reject, or escalate

One of the clearest patterns across the research was that AI made it easier to produce outputs, but not easier to know whether those outputs were actually good enough.

That distinction matters more than many teams initially realize.

AI can generate draft assessments, scripts, simulations, outlines, and summaries very quickly. But speed does not guarantee instructional quality, business relevance, regulatory safety, or pedagogical coherence. In fact, the more quickly outputs are produced, the more important evaluation becomes.

This is why judgment is becoming one of the most central capabilities in AI-enabled L&D. Professionals increasingly need to assess whether what the system produces is accurate, instructionally sound, contextually appropriate, ethically safe and sufficiently nuanced for the intended audience. In many ways, the role shifts from being the person who produces the first version to being the person who is accountable for whether the version is fit for use.

That is not a smaller responsibility. It is a higher-order one.

What strong judgment looks like in practice

This is where the human role becomes sharper, not less important.

The ability to make learning relevant to the real world of work

AI is highly effective at generating patterns. It is far less reliable at understanding the full texture of a role, culture, workflow, industry, or organizational environment unless that context is deliberately supplied.

That is why contextualization is becoming one of the most valuable human capabilities in the learning function.

Across the seven organizations, the strongest AI-enabled learning applications were not the most generic ones. They were the ones most effectively adapted to:

This was especially visible in simulation-based learning, where realism mattered deeply, and in technical or regulated settings, where generic output could quickly become misleading.

AI can generalize at scale. Humans are still much better at contextualizing for relevance, credibility, and performance impact.

Why contextualization matters more now

In practice, this means L&D professionals need to become even more attentive to the business environment, not less.

The ability to coordinate tools, workflows, review points, and learning pathways

As AI becomes embedded across the learning workflow, complexity increases.
Teams are no longer just managing content. They are increasingly managing a system of tools, prompts, content sources, review layers, governance checkpoints, learner pathways, and platform dependencies. This makes orchestration far more important than it used to be.

Across the organizations studied, the teams making the most meaningful progress were often not those with the flashiest AI tools, but those with the clearest orchestration of how those tools fit into the workflow. Someone had to decide:

This is not just coordination. It is systems-level design.

As AI capabilities expand, the learning function becomes less linear and more interconnected. Orchestration is what keeps that complexity from turning into fragmentation.

What orchestration looks like in practice

This capability is especially important for learning leaders and senior practitioners, but it increasingly matters across the entire function.

The ability to decide where automation should stop and human accountability must remain

As AI becomes more capable, the need for oversight increases rather than decreases.
This is especially true in enterprise learning, where content and learning experiences may influence compliance behavior, customer communication, technical accuracy, leadership judgment, employee decision-making, and organizational risk.

Across the cases, organizations were increasingly aware that AI could not simply be allowed to generate and deploy without guardrails. Human oversight was needed to determine where automation was appropriate, where risk was too high, and where AI outputs required review, escalation, or tighter control.

This is not only an ethics issue. It is also a performance issue.

Leaders increasingly need to ask:

These are not questions AI can answer on its own.

Why oversight is becoming more central

Oversight, then, is not a constraint on innovation. It is what makes responsible innovation sustainable.

The Human Capability Shift in AI-Enabled L&D

Many competency models in learning and development still reflect a pre-AI world. They tend to emphasize on instructional design fundamentals, facilitation skills, stakeholder management, content development, technology adoption and project coordination. These remain important. But they are no longer sufficient on their own.

As AI becomes part of the learning system, the capability profile of high-performing teams begins to expand. Future-ready competency models increasingly need to include:

The shift is subtle but important. The role of the learning professional is moving from being primarily a creator of learning outputs to being a designer, interpreter, evaluator, and steward of a more automated system.

That is a very different capability profile than many organizations are currently hiring, developing, or rewarding for.

One of the most common mistakes organizations can make at this stage is assuming that the capability challenge is mainly about tool proficiency.

Knowing how to use AI tools matters. But tool fluency alone is not enough.

A team can become highly proficient at generating AI output and still produce:

The real capability challenge is not just learning how to prompt better. It is learning how to think, evaluate, and lead more effectively inside an AI-enabled system.

That is a much higher bar, and a much more important one.

If AI is changing the capability profile of the learning function, leaders need to respond deliberately rather than assuming the shift will happen organically.

Five practical moves for leaders

This is where many organizations will either deepen their transformation or stall in superficial productivity gains.

AI is changing the capability profile of workplace learning, but not by making people less necessary. It is changing which human capabilities matter most.

As AI takes over more of the first-pass drafting, generating, summarizing, and structuring work, the value of the human contribution shifts upward toward framing, judgment, contextualization, orchestration, and oversight.

That is not a reduction of professional value. It is a redefinition of it.
The organizations that adapt best will not be the ones that simply teach their teams to use AI tools. They will be the ones that help their people become better thinkers, better evaluators, and better designers of learning systems in which AI is only one part of the equation.

The future of L&D capability will not be built on automation alone. It will be built on stronger human judgment inside increasingly intelligent systems.

Bostrom, R., & Heinen, J. (1977). MIS problems and failures: A socio-technical perspective. MIS Quarterly, 1(3).

Engeström, Y. (1987). Learning by Expanding: An Activity-Theoretical Approach to Developmental Research.

Engeström, Y. (2001). Expansive learning at work: Toward an activity theoretical reconceptualization. Journal of Education and Work, 14(1).

Leonardi, P. M. (2011). When flexible routines meet flexible technologies. MIS Quarterly, 35(1).

Orlikowski, W. J. (1992). The duality of technology: Rethinking the concept of technology in organizations. Organization Science, 3(3).

Orlikowski, W. J. (2007). Sociomaterial practices: Exploring technology at work. Organization Studies, 28(9).

Salas, E., Tannenbaum, S. I., Kraiger, K., & Smith-Jentsch, K. A. (2009). The science of training and development in organizations. Psychological Science in the Public Interest, 10(2).

—RK Prasad (@RKPrasad)