Why Your AI L&D Strategy Needs Roots First
Over the past two years, I have been in continuous conversations with CHROs, CLOs, and heads of digital learning across enterprises, digital publishers, and learning technology platforms. Almost every organization has an AI learning initiative in motion. Investments are being made, pilots are underway, and expectations from the C-suite are high.
But when these conversations move past the surface, a consistent and uncomfortable pattern emerges.
Despite record levels of AI investment in L&D, measurable impact on workforce performance remains elusive. Content is being produced faster, but not applied better. Pilots are visible on dashboards but aren’t scaling. And skills gaps, the ones executives most urgently want closed, remain stubbornly wide.
According to BCG, 74% of organizations report no tangible business value from their AI investments, despite a collective $252.3 billion in AI spending in 2024 alone. MIT’s 2025 GenAI in Business study found that 95% of GenAI pilots fail to demonstrate P&L impact, and S&P Global reported that 42% of companies abandoned most of their AI initiatives in 2025—up sharply from just 17% the prior year.
In learning specifically, LinkedIn’s 2025 Workplace Learning Report flags that while 80% of L&D professionals view AI as important to their strategy, only 25% factor it into their work routinely. Meanwhile, 49% of learning and talent professionals say their executives are concerned that employees don’t have the right skills to execute business strategy.
This is the AI learning gap no one is talking about loudly enough: the gap between investment and real workforce capability.
In my view, and in Harbinger’s work supporting some of the world’s leading digital publishers, associations, and enterprise learning teams, the root cause is not the technology. It is the foundation on which AI learning strategies are being built.
AI In Learning Is Not A Tool Upgrade But A System Shift
The most common starting point I see is organizations treating AI as a faster way to do what they were already doing: build courses more quickly, generate assessments at scale, or automate translation and localization. These are real efficiencies. But they don’t change how learning operates.
AI fundamentally changes the economics of learning content. What used to take 40 hours now takes 4. But if the content still sits in SCORM packages that no one opens past slide 12, you’ve just produced mediocrity faster. Learner expectations are also shifting: people want support embedded in the flow of work, contextual and just-in-time, not a course launched from an LMS.
This creates a structural demand on the learning ecosystem that most organizations are not yet meeting. Content can no longer be static. Systems must evolve continuously. The underlying architecture must support modular reuse, AI interaction, and contextual delivery across channels.
When organizations layer AI onto legacy, course-centric models without addressing these structural realities, the results are predictable. AI doesn’t transform a broken system. It exposes and accelerates its limitations.
Where Most AI Learning Strategies Break Down
Across enterprise engagements and digital publishing transformations, Harbinger has consistently seen the same failure patterns.
Content unreadiness: Most learning ecosystems are built on SCORM packages, PDFs, and linear video—formats designed for delivery, not for machine interaction. Without structured metadata and modular architecture, AI systems lack the context needed to generate reliable outputs. The result: more time spent validating AI-generated content than benefiting from it.
McKinsey’s 2025 State of AI report highlights that 51% of organizations experienced at least one negative AI-related incident in the past year—most commonly output inaccuracy and compliance violations—a significant liability in regulated sectors.
Treating modernization as a one-time project: Organizations launch a content migration or a platform upgrade and then wait for the next budget cycle. In an AI-driven environment, content cannot remain static. Without continuous modernization workflows, organizations find themselves perpetually behind.
Governance as an afterthought: AI enables speed. But without embedded governance, that speed introduces risk. Organizations frequently hesitate to scale AI because they lack confidence in how errors will be detected, corrected, and audited.
Role ambiguity inside the learning function: As AI enters workflows, instructional designers, SMEs, and QA teams are often unclear about how their work evolves. This ambiguity creates friction and slows adoption not because people resist AI, but because no one has redesigned the operating model.
Disconnection from business outcomes: Perhaps the most critical failure. Most AI learning strategies are measured in efficiency terms, like hours saved and courses produced. Business leaders are now asking a different question: are our people actually more capable? Are we closing the skills gaps that matter? When learning remains centered on content production rather than capability building, it struggles to answer that question honestly.
What The Evidence Shows About High-Maturity Organizations
LinkedIn’s 2025 Workplace Learning Report is instructive. Only 36% of organizations qualify as “career development champions”: those that systematically connect learning to career pathways, internal mobility, and business outcomes. But those that do see measurably different results: higher profitability, better talent retention, and significantly stronger AI adoption rates. Career development champions are 32% more likely to offer AI training and 51% more likely to consider themselves frontrunners in generative AI adoption versus just 36% for less mature organizations.
The pattern is consistent with what we see in Harbinger’s own delivery work: the organizations that see the most from AI are not the ones that started earliest with the tools. They’re the ones that first got their content infrastructure and operating model right.
Illustrating with two examples from our work.
In one large-scale course industrialization engagement—similar to work done with healthcare and compliance content publishers—an organization had thousands of courses, each customized for different audiences. Rather than migrating content as-is, the decision was made to restructure it into reusable learning objects with proper metadata tagging. What followed was a 10x increase in content production speed and an 80% automation rate, but more importantly, the modular structure meant content could be updated once and republished across formats automatically. AI was the accelerant; the architecture was the foundation. (This mirrors work we have done for clients in the healthcare and compliance training space, including a 6000-course automation initiative in the clinical education sector.)
In another case, a leadership development organization moved from static course formats to a structured, single-source content model. Once content was modular and metadata-rich, AI-powered personalization became viable not because they adopted a new tool, but because the content was finally machine-readable. AI coaching simulations, dynamic assessments, and adaptive pathways all became possible as downstream applications of structural work that was done first.
The pattern: system design precedes AI value capture.
A Practical Model: Content Maturity × Operating Model Maturity
It helps to think about AI learning strategy across two dimensions: content maturity (how structured, modular, and reusable the content is) and operating model maturity (whether the learning function runs on project-based workflows or continuous delivery).
Organizations with unstructured content and project-based workflows find that AI creates more rework than value.
As content becomes more structured, reuse and consistency improve…but without operating model changes, scale stays limited. True transformation happens when both dimensions mature together. High-maturity organizations build modular content systems supported by continuous workflows and embedded governance. In these environments, AI becomes a natural system extension rather than a bolt-on.
This dual-maturity lens is how Harbinger approaches AI readiness conversations with clients, whether they are enterprise L&D teams trying to move from content delivery to workforce capability or digital publishers trying to transform a catalog of PDFs into an AI-ready content supply chain.
What High-Maturity Teams Do Differently
The most sophisticated learning organizations I have worked with share a defining characteristic: they don’t begin their AI journey with tools. They begin with system design.
They treat content as infrastructure, not as finished product. Content is broken into modular components, enriched with metadata, and designed for reuse. Courses, performance support tools, AI copilots, and analytics systems can all draw from the same source.
They rethink assessments. Instead of fixed, linear assessments embedded in courses, they build dynamic systems where questions are tagged by skill, complexity, and context. This allows assessments to adapt based on learner responses and generates richer data about actual capability development, not just completion.
They redesign roles, not just retool them. Instructional Designers become experience architects. SMEs shift from content producers to knowledge validators. QA expands into AI governance not as a bottleneck, but as an embedded quality and compliance function. This is the workforce transformation piece that most AI learning strategies miss entirely.
They embed governance from the start. High-maturity organizations define clear boundaries for where AI can be generative and where it must remain deterministic. Audit trails and traceability ensure that innovation does not compromise trust—especially critical in regulated industries.
And they measure differently. Rather than tracking content volumes or completion rates, they track skill progression, internal mobility, and performance improvement. They answer the question that matters to business leaders: are our people becoming more capable at the things that drive business outcomes?
Where To Start
For organizations looking to strengthen their AI learning strategy, the starting point is not a new tool or a new platform. It is an honest diagnostic.
Three questions worth asking:
- Is your content structured in a way that supports modular reuse and AI interaction, or is it locked in formats designed for one-time delivery?
- Are your learning workflows designed for continuous evolution, or do you operate on budget cycles and project timelines that make ongoing improvement structurally difficult?
- Is governance embedded into how AI is used in your content supply chain, or is it applied after the fact, creating the hesitation that prevents scaling?
Answering these honestly provides a clearer roadmap than any technology evaluation. For organizations that want a structured benchmark, Harbinger’s CLEAR Content Audit Framework provides a scored diagnostic across content quality, AI readiness, learner experience, and library rationalization.
Closing Thoughts
The future of learning is not defined by how fast content can be created. It is defined by how effectively organizations can build systems that develop real workforce capability continuously, at scale, and in alignment with where the business is going.
At Harbinger, we work at the intersection of digital publishing, workforce enablement, and talent transformation. What we consistently find is that the organizations making the most of AI in learning share one thing in common: they invested in the foundation before they invested in the features.
AI is a powerful enabler of workforce transformation.
But only when the system is ready to receive it.
