During the first .com boom, businesses scrambled to understand the internet’s potential. Amazon famously took nearly a decade to become profitable, and Jeff Bezos tirelessly explained to investors why change was necessary while clinging to his vision of the future. Today, learning and development (L&D) faces a similar challenge — the challenge of producing now while planning and preparing for what’s to come.

L&D teams are on a journey to meet the expectation that they use artificial intelligence (AI) to deliver faster, better and more effective learning. At the same time, learning leaders are preparing for a future of learning that looks radically different than today’s AI strategy, which often means simply layering AI use onto legacy L&D practices. How do L&D teams and learning leaders balance these competing demands now? Can we meet today’s demands while preparing for tomorrow’s?

The AI Maturity Model for L&D

Learning design teams need to be comfortable with AI. They need clear usage guidelines, shared resources and functional, collaborative workflows. At the same time, learning leaders need to be preparing their teams for transformation.

To illustrate the range of AI maturity and adoption, we created the following five-phased maturity model based on our experience with clients and the broader industry. While other AI maturity models exist, our goal was to create something practical that can serve as a roadmap and provide an intentional and structured way to assess 1) where your team is, and 2) where your team needs to go.

• Phase 1. Ad Hoc: Individuals use AI sporadically and independently. There is no formal process, governance or shared knowledge across teams or the business. Individuals may have a foundational understanding of AI, but its adoption is generally low, and experimentation is informal.
• Phase 2. Exploratory: Small groups of individuals explore AI together, and they are in the early phases of sharing best practices and templates. Foundational practices like role clarity and workflow are beginning to form. The AI capabilities these teams are exploring are usually role specific.
• Phase 3. Structured: Teams adopt standardized processes for AI use. Governance, version control and ethical guidelines are established. Teams select AI tools tailored to their needs and begin making AI-informed decisions.
• Phase 4. Integrated: The organization begins to intentionally embed AI into core workflows across multiple teams. The organization is beginning to integrate AI cross-functionally across L&D, human resources (HR), compliance and IT in order to manage and optimize processes.
• Phase 5. Transformational: The organization is purposefully leveraging AI to drive strategic transformation. Enterprise-wide adoption, continuous improvement and measurable impact are now the norm. Human-AI collaboration is also normalized, and responsible AI use has become a core value.

Transformational AI use will allow organizations to become truly learner-centric by making them more agile in the face of change, more connected and more collaborative.

Ultimately, this will enable the learning industry to move from being rigid and course-centric to providing layered learning experiences that evolve over time based on individual learner needs. This means that, in Phase 5, learning practitioners will have shifted from creating static content to enabling adaptive learning ecosystems through atomic instructional design — something that, while seemingly futuristic, is becoming increasingly real.

Where Learning Teams Are Today: Embedding AI Into Workflows and Processes

While the model outlines the full spectrum of maturity, our work in the industry has shown us that most teams are in Phase 2 or early in Phase 3. Most AI use is still role-specific, but structure around guidelines is beginning to form and tools are being shared.

The priority for organizations in these early phases lies in embedding AI into L&D workflows, which requires process integration. For this integration to happen and to prepare for future transformation, learning teams need to operate with clarity and structure, supported by both traditional project management practices and AI-enhanced capabilities.

For your team to become less exploratory and more structured in its AI use, start with these foundational practices:

Centralize AI Resources

Why? Building a central repository for prompts and agents enables teams to share, refine and reuse best-practice queries and engineered prompts. This improves consistency, reduces duplication and accelerates development, making it easier for teams to collaborate and maintain quality across AI-driven projects.

How? Treat prompts and agents as deliverables in their own right, analogous to tools and templates in an engineering workshop, and manage them with the same attention to tolerance, wear and tear and entropy. Adopt version control for engineered prompts and agents, ensuring they are accessible and updatable by the team.

Implement Robust Version Control for Prompts

Why? Documenting changes to prompts ensures auditability, supports refinement and makes knowledge sharing easier. Version control also helps teams revert to previous states when needed, reducing risk and improving transparency.

How? For saving multiple versions of prompts, use tools that provide the ability to go back to a previous version, ensuring that the workspace or components can be worked on by multiple roles with visibility of contributions.

Enable Multi-Role Collaboration

Why? This fosters accountability and enhances teamwork, ensuring that edits and decisions are traceable and aligned with project goals.

How? Select AI tools that support collaborative design and allow for parallel projects involving both internal and client personnel. Use shared workspaces where designers and contributors can work together on outputs, with visibility into who has contributed and made changes. Ensure the workspace/components can be worked on by multiple roles, similar to how a Word document accepts tracked changes, providing transparency and traceability.

Invest in Data Provenance

Why? Data provenance practices ensure that every piece of generated content can be traced back to its origin, which is critical in regulated environments and for confidential material. This practice supports compliance, builds trust and simplifies audits.

How? Include references to where combined data has come from (source acknowledgements and tracking). Validate the provenance of additions to the dataset, ensuring that all inputs are managed and traceable.

Implement Retrieval-Augmented Generation (RAG)

Why? This approach reduces manual fact-checking and ensures AI-generated content reflects the latest authoritative information.

How? To use RAG, specify the exact data sources within the larger dataset for the AI to reference, such as internal policy documents, compliance checklists or trusted databases. Provide the AI tool with relevant documents and checklists so it can generate content that is accurate, up-to-date and aligned with internal standards.

RAG can be done either by user intervention or by using an embedded large language model (LLM) that searches and ranks documents within the dataset.

What Will Future Learning Look Like? Performance Enablement.

Once foundational practices are in place, the next horizon is performance enablement, where learning becomes dynamic and embedded in work. At this point, learning design teams will need to shift from creating static content to building adaptive ecosystems where learners influence outcomes. Objectives will evolve dynamically, guided by context and supported by AI-driven personalization.

While many learning teams are occupied with structured AI use (embedding it into workflows and processes), they must also be preparing for this future that will shift their role from delivering content to enabling performance.

Here’s how your team can start building the mindset needed for the future of learning:

• Embed learning in the flow of work. Integrate development opportunities directly into employees’ daily workflows. When learning happens in real time, it feels practical and immediately applicable, driving productivity, adaptability and continuous skill growth — cornerstones of performance enablement.
• Personalize learning. Design experiences that reflect individual career aspirations and encourage cross-functional exposure. Tailored learning boosts engagement, strengthens internal capabilities and supports talent retention. Personalized, purpose-driven development is essential for cultivating future leaders.
• Build future-ready skills. Focus on capabilities like adaptability, strategic thinking, digital fluency, data analytics and AI literacy. Prioritizing these areas enables L&D teams to craft data-informed, personalized learning experiences that move beyond content delivery to truly empower performance.
• Become a strategic business partner. Ensure every learning initiative is tied to organizational objectives. This alignment turns L&D investments into strategic levers that deliver measurable impact and reinforce enterprise goals.

We have been trying to solve these problems for years, so they are not new considerations in our industry by any stretch. The difference is that now, with AI, we can actually deliver on them.

Final Thoughts: Embrace the Disruption or Be Disrupted

AI is changing the way learning happens and L&D teams need a clear path to adapt. Our AI maturity model provides that roadmap, guiding teams from early experimentation to full integration and strategic impact. The practical steps outlined earlier are essential building blocks. These actions help teams establish structure, improve efficiency and prepare for advanced phases where AI becomes part of everyday workflows and redefines learning as we know it.

Learning teams have two choices: embrace and harness the disruption or be disrupted. Organizations that start now will be ready to lead in the future.