What is an AI upskilling program?

AI upskilling programs are designed to help employees develop the skills needed to work with AI in their current roles. They focus on building job-relevant AI capabilities and preparing your team for the future of AI. 

An AI upskilling program may center on how employees interact with AI as part of their responsibilities. This could include when to use it and how to apply AI results in context, with the goal to support consistent, role-appropriate use.

As an employer, you may choose to formalize AI upskilling so expectations are clear as AI becomes part of everyday work. A defined program can help you clarify how AI-related skills fit into existing roles today and how those expectations may need to change as work shifts over time.

Before you start: principles to guide AI upskilling

AI upskilling can take different forms depending on your organization. It can help to keep a few guiding principles in mind that shape how your employee upskilling effort is planned and applied.

Focus on real work and outcomes

You may want to think about AI skills in the context of everyday responsibilities and expected results. Anchoring skills to real work can make conversations about AI more concrete and easier to apply.

Emphasize observable, job-related skills

Defining AI skills in terms of actions and decisions employees make at work may reduce ambiguity. Doing so may help align expectations consistently, rather than being interpreted differently across teams.

Allow space for practice and feedback

AI-related skills may develop through use over time. Creating room for application and review can help learning connect to work situations.

Plan for ongoing adjustment

AI upskilling may shift as roles, needs and technology change. Treating any reskilling and upskilling efforts as adaptable processes may make it easier to adjust direction as conditions evolve.

Steps to build an AI upskilling program

The following steps outline a practical approach to building an AI upskilling program.

Step 1: Align AI upskilling to business priorities

Before focusing on skills or learning activities, it can help to clarify what the upskilling effort is meant to support. Connecting AI upskilling to specific business priorities makes the program easier to scope, evaluate and integrate.

One way to approach this step is to think about the outcomes you want AI-related skills to support. Depending on your organization, this might relate to how work is performed in certain roles, how teams manage workload or how you plan for future talent needs. Framing AI upskilling around these kinds of goals can provide direction while leaving room to adapt.

To clarify those priorities, you might consider:

• Which parts of the business could benefit from changes in how work is done, rather than additional headcount
• Where roles are expected to change in the near or medium term
• Which teams or functions would benefit most from stronger AI-related skills today
• How success is described in business terms rather than training activity

When AI upskilling is treated as a business effort rather than a training exercise, it may be easier to align expectations across teams. It may also help keep the focus on how AI skills relate to day-to-day work and longer-term planning.

Step 2: Identify roles and levels for AI upskilling

AI skills don’t need to be the same across every role. Different jobs may interact with AI in different ways, which can influence the type and level of skills that are most relevant.

For some roles, AI skills may involve understanding AI outputs or using AI-supported processes. For others, the focus may be on reviewing results, making decisions or shaping how AI is applied within the role. Thinking in levels can help keep expectations realistic and aligned with job needs.

To make this more practical, you might consider:

• Which roles are most likely to interact with AI as part of their regular responsibilities
• Where AI directly impacts decision-making, quality or speed
• Which roles may need a general understanding versus a deeper, role-specific capability
• How role groupings today could change as responsibilities shift over time

This kind of role-based view can support a more targeted AI reskilling strategy by helping you focus your efforts where they are most relevant. It also allows for flexibility as roles evolve, without requiring every employee to follow the same upskilling path.

Step 3: Assess current AI skills across your workforce

Understanding where your employees are starting from can help you scope your AI upskilling effort and set realistic expectations.

At this stage, you might search for simple indicators of current capability rather than formal assessments. These indicators describe how AI appears in everyday work and how your employees relate to it. They could include:

• Knowledge: familiarity with AI-related concepts or terminology relevant to the role
• Confidence: comfort level when discussing or engaging with AI-supported tasks
• Usage: whether and how AI is part of regular work activities
• Hesitation: reluctance to use AI because of uncertainty, lack of clarity or concerns about impact on work

To gauge these indicators, consider using a short, role-aware survey or check-in. This could include a small set of questions that ask your team how they currently interact with AI, how comfortable they feel and where questions or hesitation exist. Input from your managers or team-level discussions may also help add context, especially when usage varies within the same role.

Step 4: Define the AI skills and competencies to build

With an outline of roles and current skill levels, you can begin clarifying which AI-related skills are relevant across your workforce. Clear definitions can help managers and teams align on expectations and reduce ambiguity as AI becomes part of everyday work.

One way to keep skill definitions practical is to translate broad AI-related needs into observable, job-related actions. This helps translate general intent into expectations that employees can apply in their roles.

For example, broad AI-related needs can be expressed in more concrete terms:

As you define skills, you may want to consider:

• Using language tied directly to job roles and responsibilities
• Focusing on actions or decisions that employees can demonstrate in their work
• Avoiding abstract labels that can be interpreted differently across teams

These definitions can then serve as a foundation for learning paths and practice opportunities at later stages of your plan.

Step 5: Create a role-based AI upskilling plan

Once you’ve identified priority roles (Step 2) and defined the AI skills to build (Step 4), the next step is to turn those definitions into a plan employees can follow.

You can approach this step role by role, starting with one role group and expanding over time.

1. Link each AI skill to a specific work activity

Begin by identifying where each AI skill appears in day-to-day work. This helps move from abstract skill definitions to practical application.

For each role, ask:

• Which tasks or decisions would involve this AI skill?
• Where would stronger AI use change how work is performed, reviewed or completed?

The goal is to clarify where AI fits or is likely to fit within existing responsibilities.

2. Decide how employees will build the skill

Next, determine what support employees need in order to use each skill effectively.

Depending on the skill and role, this might include:

• Short explanations or examples of effective AI use
• Clear guidelines on when and how to apply AI
• Role-specific prompts, templates or checklists
• Shared examples of AI outputs and how they were evaluated

Keeping learning inputs lightweight can make it easier to integrate skill-building into ongoing work.

3. Plan for on-the-job practice

AI-related skills develop through use. As part of the plan, identify how employees will practice each skill in real work situations.

You might consider:

• Which tasks will employees apply the skill to
• How often are they expected to use AI as part of that work
• What “good use” means in context, based on the job-related skill definitions you established in Step 4

This step helps reinforce AI upskilling through everyday responsibilities rather than being limited to one-time learning activities.

4. Identify feedback and support points

Finally, consider how employees will receive feedback as they build these skills. Feedback helps reinforce expectations and surface questions or issues early.

Support may include:

• Manager check-ins focused on how AI is being used in work
• Team discussions to review AI-supported outputs
• Shared reflection on what worked well and what needs adjustment

These feedback points help connect learning to performance and allow the plan to evolve over time.

Basic example of a completed AI upskilling plan

To keep planning manageable, you may find it helpful to capture this information in a simple format. The following example illustrates how AI skills can be linked to real work, learning support and practice for a single role.

Example: Role-based AI upskilling plan for a marketing manager

While simplified, this structure can serve as a practical AI training framework that you can reuse across roles and update as needs change.

Step 6: Enable on-the-job practice and adoption

After creating your plan, the next step is to embed AI use into how work is reviewed, discussed and evaluated. At this point, the focus shifts from planning practice to reinforcing expectations so AI use becomes part of normal work routines.

You may want to find where work is planned and reviewed, and bring AI expectations into those moments. This can include project kickoffs, regular check-ins, team reviews or manager feedback conversations.

Referencing the AI skills and work activities defined in your plan during these discussions helps make AI use visible and expected.

Adoption is reinforced when AI-related decisions are treated like other work decisions. Asking when AI was used, how outputs were reviewed and why certain results were accepted or adjusted helps normalize these behaviors without adding extra steps or formal employee training.

You could support this shift by:

• Incorporating AI-related questions into existing review or check-in conversations
• Expecting AI-supported work where your plan indicates it is relevant
• Using the job-related AI skill definitions from Step 4 when discussing performance or quality

For example:

For a team producing a recurring report, managers can routinely ask how AI was used, how outputs were verified and how final decisions were made. Over time, these questions become a standard part of the review process.

Step 7: Measure outcomes and refine your AI reskilling strategy

Once AI skills are being used in real work, the final step is to check where the upskilling is making a difference and where it needs adjustment.

What to measure

You might check:

• Whether AI skills are being used in the tasks identified in your role-based plans
• Whether AI outputs are being reviewed before use
• Where AI use is improving the quality, speed or consistency of work

How to measure consistently

To keep measurement simple and comparable across teams, choose one consistent approach and use it over time. This could include:

• Repeating the same short survey used in Step 3
• Asking managers the same set of questions across teams
• Reviewing a small sample of completed work using the same skill definitions

Use what you learn to make small, targeted changes. This may include refining skill definitions, updating role-based plans or expanding upskilling to additional roles where it’s working well.

Common challenges when building an AI upskilling program

Even with a clear plan, you may encounter challenges when upskilling your workforce. These may include:

Lack of clarity on goals

When AI upskilling is not clearly tied to business outcomes, it can be difficult for managers and employees to understand why certain skills matter or how to evaluate success. This may lead to uneven adoption or confusion about priorities.

One-size-fits-all training approaches

When AI upskilling is delivered as a single program for all roles, it can feel disconnected from day-to-day work. Research suggests that a meaningful share of workers do not currently use AI or find a need for AI-related training, which may limit engagement if expectations are not role-relevant. Aligning training to job responsibilities may help address this gap.

Limited opportunities for practice

AI skills develop through use. Without clear opportunities to apply AI to regular tasks, learning remains theoretical and confidence stalls.

Difficulty measuring impact

Early on, it may be challenging to connect AI upskilling efforts to visible outcomes. Without clear signals tied to work quality or decision-making, it may be harder to assess progress or know where adjustments are needed.

AI upskilling is an ongoing process that evolves as roles, tools and business needs change. Starting with a clear, structured approach can help you set expectations and adapt over time without overcomplicating the work.