How Learning at Work is Changing with GenAI

Learning is a fundamental human desire. We're naturally curious, driven to grow and to understand more about the world around us. This is especially true in the workplace, where learning isn't just about getting the grade or checking a box. Rather it’s about staying relevant, getting promoted, improving processes, or simply doing our jobs better.

*Image from Unsplash. I love learning stuffs.*

But here's the thing: corporate learning has been stuck in a rut for years. Traditional training departments became content factories, churning out videos and workshops that rarely translated into real skill development. That's starting to change now, thanks to Generative AI (GenAI).

Why Corporate Learning Hit a Wall

Over the past decade, Learning & Development departments transformed into publishing houses. They focused on creating training materials and designing courses, but lost sight of their real purpose: driving organizational change and transformation.

The results speak for themselves. Despite billions invested in corporate training, we keep seeing the same issues: poor engagement, minimal skill transfer, and leaders who couldn't tell if their training programs actually worked. Just because you create an hour-long video or run a day-long workshop doesn't mean people become better managers, leaders, or performers.

The traditional model is somewhat broken, and stale.

Why GenAI Changes Learning

GenAI is uniquely suited for learning in ways that previous technologies weren't. It can absorb massive amounts of content and repackage it into virtually any format. It can organize information, chunk it appropriately, and make it engaging through conversation.

Josh Bersin, who has been analyzing the corporate learning market for decades, puts it perfectly:

"AI is the most perfect technology ever created for learning. All the things that we've tried to do by hand in the publishing model of training can be done by AI."
Source: How AI Is Blowing Up The Corporate Learning Market: The Whole Story – JOSH BERSIN

He describes GenAI as a "dynamic content system." Unlike static courses or videos, GenAI can adapt content in real-time based on individual needs, learning styles, and progress.

GenAI can handle the content creation piece better than humans in most cases. It has access to vast amounts of information, case studies, and research, and can create outputs in any format: audio, video, interactive chatbots, checklists, or guides. Yes, it makes mistakes, but so do humans. And when GenAI costs 10% of what human content creation costs, the choice becomes obvious for most organizations.

Beyond content creation, GenAI will excel at micro-learning in context, reviewing and refreshing content, analytics, and compliance tracking. It can monitor learning progress, identify gaps, and suggest interventions in ways that human administrators simply can't match at scale.

The Rise of Learning Agents

We're already seeing GenAI tools like ChatGPT and Microsoft Copilot being used as learning assistants. These "open agents" work across different interfaces: voice agents on your phone while commuting, desktop agents for text and video, and someday virtual reality agents for immersive learning experiences.

But the real transformation will come from custom agents designed for specific organizational needs:

Research Agents will gather and synthesize learning content from multiple sources, ensuring it's current and relevant.
Update Agents will continuously refresh existing materials, keeping pace with changing regulations, best practices, and industry developments.
Compliance Agents will monitor completion rates, assess learning effectiveness, and ensure regulatory requirements are met.
Onboarding Agents will provide personalized guidance for new hires, adapting to their role, experience level, and learning pace.

What Learning Designers Do Next

In the short term, learning professionals will become curators and editors for GenAI. They'll be the quality gatekeepers, ensuring AI output meets organizational standards and truly serves learners' needs.

But there's a crucial limitation to understand: GenAI can't teach everything. Some skills require human coaching, especially those involving:

Complex physical or emotional components (like learning to swim or handling difficult conversations)
Organization-specific processes and systems (your customized SharePoint setup, unique workflows, cultural exceptions)
Institutional knowledge transfer (the "how we really do things here" versus official procedures)
Context-dependent problem-solving (navigating company politics, understanding stakeholder relationships)

A Real-World Example

*Image from Unsplash. Who loves to swim?*

Take learning to swim as an adult, which I did a few years ago. Here's what that process actually looked like:

First, I had to face my fear of water. My first session with a coach was mostly talking, not swimming. I needed to work through the psychological barrier of being submerged, learning to breathe out underwater. It was incredibly frustrating, almost to the point of tears.
The next few sessions involved a lot of resistance and arguing with my coach. I'd refuse to go into the deep end, get frustrated, and need calming down. My coach had to read my emotional state and know when to push and when to give me space.
Then one day, something clicked. I found myself swimming in the deep end, slowly but surely.
Later sessions focused on technique refinement: hand placement, leg movement, stroke efficiency. Things I couldn't observe about myself while in the water.

Throughout this process, my coach provided something GenAI couldn't: real-time emotional support, personalized encouragement, and the ability to adapt instruction based on my psychological state, not just my technical progress.

Could GenAI have helped me learn to swim? Absolutely. It could have provided theory, tips for overcoming fear, and technique guidance. But the practice, the real-world application, the moments of breakthrough and frustration, those required human insight and support.

How this Applies to Organizational Learning

Context-specific, process-integrated learning is where humans truly shine. GenAI can explain how SharePoint works in general, but it can't say (without special knowledge):

"Here's why we set up our document library this way because of how our legal team reviews contracts."
or
"This is the workaround we use because our IT department configured permissions for external users differently."

This is about the kind of learning that's embedded in how a specific organization actually operates:

Organizational workflows and exceptions ("We don't use the standard approval process for these types of requests because...")
System customizations and workarounds (“We apply permissions on this site differently because external users are invited.” )
Cultural context ("The CEO always wants to see this data formatted this way")
Relationship dynamics ("You'll need to loop in Sarah from accounting because she handles the exceptions")

This strengthens the case for human instructors long-term, but in a different way than traditional training. Instead of being course creators, they'll be more like organizational anthropologists — people who understand how work actually gets done (versus how it's supposed to get done) and can help others navigate those realities.

This kind of learning is also inherently social and collaborative. A SharePoint refresher involves people sharing their own tips, asking "what do you do when..." questions, and learning from each other's mistakes. That peer-to-peer element is hard for GenAI to facilitate authentically.

The Future of Learning: Human + GenAI

This is how I see learning evolving: instead of front-loading theory through lectures and videos, we'll flip the model entirely.

Learners will use GenAI tools to absorb foundational knowledge at their own pace. Then they'll spend their time with human instructors who provide real-time feedback during practice and application.

In schools, this might mean students learn concepts at home with GenAI tutors, then use classroom time for active practice: writing, problem-solving, and discussion, with teachers providing immediate, personalized feedback.

In corporate settings, instead of sitting through presentations or watching training videos, people will work alongside mentors and senior colleagues who coach them through real situations. They'll attend actual meetings, handle real customer interactions, and tackle genuine business challenges, with experienced professionals providing guidance and feedback in the moment.

But here's what organizations are missing: Learning isn't a one-time event. The traditional "one-and-done" training model fails because:

New team members join and need to learn systems and processes
People forget details about features they don't use regularly
Systems evolve with updates and new capabilities that require fresh learning
Organizational processes change and require refresher training

The solution is ongoing learning support rather than project-based training. Think monthly office hours, quarterly refreshers, and on-demand mini-sessions for new features or team members. GenAI can make this ongoing learning program more efficient.

This approach makes learning more human, more immediate, and more actionable. It pairs the efficiency of GenAI with the value of human wisdom and emotional intelligence.

Next steps to get started

If you’re leading learning and organization change, don’t make the biggest mistake I see: Treating learning like a project instead of an ongoing process.

Use GenAI to invest in initial training but also plan to budget for:

Refresher training for existing team members
Onboarding support for new hires
Feature updates when systems evolve
Context-specific guidance that GenAI can't provide

The organizations that win will shift from one-time training purchases to ongoing learning partnerships. They'll recognize that institutional knowledge transfer and system-specific guidance require human expertise that complements GenAI's capabilities.