How to Analyze Employees’ AI Skills Gaps

When we hear the term “artificial intelligence (AI) skills,” many of us picture things that feel far from the workplace we know. 

Data science teams rattling off code. Programming glowing-eyed robots to spin all the plates we can’t. Building super-computers capable of distilling the meaning of life before we’ve had the first sip of our morning coffee.

In the future of work, these are the skills we think we’re supposed to need. But in reality, the AI skills organizations need most today are much more grounded — and well within reach. 

In a skills landscape that’s continually shifting, the real challenge isn’t just knowing which AI skills exist — it’s knowing which ones matter most to your employees and business goals. To do that, you need to know which ones you’re missing.

What do we mean by AI skills?

AI skills refer to the subset of technical and non-technical abilities employees need to develop to get the best from any AI-enabled tooling. 

What that looks like depends on the team. Your engineering department might be dabbling in machine learning and natural language processing, while marketing is refining prompts to improve campaign performance. 

But for most employees, high-demand technical skills include:

• AI awareness: Understanding the different types of AI models and tools, how they work, what each can and can’t do, how to assess which one is right for your role, and understanding where tools might fail. 
• Data literacy: Understanding what data you have, how to analyze and interpret it, and how to use it to inform decisions — including AI-generated insights. At a more advanced level, this could include how data is structured and can be used to build AI workflows, training models, automations, agents, or custom GPTs.
• Advanced analytics: Using AI to analyze patterns, predict outcomes, and generate insights from complex datasets.
• Prompt engineering: Writing clear, specific prompts that generate useful, accurate outputs — and learning how to refine them when they don’t.
• Workflow automation and custom tools: Building workflows and automations to improve processes and solve team-specific problems.

But for AI to embed long-term, organizations must couple technical skills with human ones — the ones that a computer can’t replicate.

“To build a career with longevity, employees need to double down on the human side,” said Matt Bradburn, former people leader and founder of AI-HR consultancy PeoplexAI. “The three key areas teams need to develop are curiosity (what we do), commerciality (what should we do), and critical thinking (what is the best way to do it) — so you’re solving problems that impact the business.”

In the AI-powered workplace, the ability to reason, make decisions, critically evaluate outputs, and communicate will help employees challenge and contextualize what AI produces. Meanwhile, a strong foundation in ethics and responsible use — including bias awareness, data privacy, and security — means that AI will be used with care.

Honing both the technical and non-technical skills simultaneously is where organizations will move beyond experimentation into true fluency — and where AI becomes a driver of business value.

The Business Case for Closing AI Skills Gaps

For many organizations, the swell of AI has felt less like a strategic rollout and more like they’ve been dropped into a whitewater rafting expedition — without the raft.

As a result, employees have largely led the charge for change. According to a 2024 report by Microsoft and LinkedIn, 75% of surveyed knowledge workers are using AI-powered tools at work, while 78% of AI users are bringing in their own AI tools to get work done. 

This has led to a ground-up, uneven approach to adoption. Employees are developing skills in individual silos, often without guidance — and those skills don’t always align with what the business actually needs. 

But if organizations take the lead in setting clear policies and upskilling pathways to address skills gaps in line with their business goals, they have a real opportunity to turn this fragmented AI usage into focused capability. 

In the near term, this means improved productivity, clearer compliance, and better risk mitigation. 

But in the long term, it means building a skills-based competitive advantage — increasing your ability to attract top talent, engage employees with meaningful growth paths, and drive innovation. 

How to Identify Skills Gaps in Your Workforce

Your organization’s biggest AI skills gaps won’t present themselves neatly in a map or survey. They show up in the messy parts of work and day-to-day processes — missed performance targets, clunky team collaboration, and teams feeling the pressure to implement tools they don’t fully understand.

Some of these signals are measurable, and some aren’t. Which is why identifying your AI skills gaps is less of a metric hunt, and more a case of building an ugly facial composite by following a bunch of scattered clues. 

To follow this breadcrumb trail, you need to look through multiple lenses to identify how your employees and teams feel, what your systems show, and how market shifts are evolving.

1. Evaluate your employees’ level of understanding and comfort toward AI.

While some employees might have already rolled out the red carpet for AI and become overnight super-users, others may be highly suspicious, struggling, or downright scared that their job is just one automation away from being squished out of existence. 

Acknowledging these human fears and knowing what’s standing in the way can help organizations scale adoption and confidence faster.

“This is the point where I’d do a listening tour,” said Theresa Fesinstine, founder of AI consultancy peoplepower.ai and author of People Powered by AI: A Playbook for HR Leaders Ready to Shape the New World of Work.

“I’d run conversations with employees to ask questions about AI usage so that people don’t feel they have to do it in secret or use tools the company wouldn’t support. It’s focused on building that educational piece — how can we actually use these tools to our greatest capacity instead of it just being a new, shiny object?”

Running pulse surveys or including AI-related questions as part of regular engagement surveys can help organizations track changing sentiment and AI expertise over time. Fesinstine suggests using the following questions to gauge confidence levels and identify blockers:

• Have you started to explore AI technologies in your work yet?
• What’s that experience been like? What has AI been most helpful for?
• If you’re not finding AI helpful, can you share how you’ve been using it?

{{rich-highlight-1}}

2. Assess your current skills landscape at a team level.

AI confidence starts at an individual level. But capability (or a lack of it) shows up in teams and how employees get work done on a day-to-day basis. One team might be deep into building algorithms and launching new pilots to optimize routine tasks, while others are just using generative AI to finesse the odd email or social media post.

This kind of variability — in both adoption and use cases — means you’re not just looking for measurable outcomes, like performance metrics. You’re also looking for the intangible characteristics that drive long-term fluency.

“AI tools change every month, which makes it hard to quantify the skills needed to use them,” Bradburn said. “The thing that matters most is that people can stay adaptable, be curious, and keep learning.”

This is where always-on performance management processes become especially helpful. 

Regular touchpoints like one-on-ones, peer reviews, and manager evaluations provide teams with two types of data:

• Performance signals: Which employees are achieving their goals? Who are our highest performers? What are employees’ key opportunities for growth?
• Growth signals: Are employees trying new approaches without being prompted? What use cases are employees bringing up in team conversations? 

When viewed together, these data points help teams spot both the visible gaps and the invisible ones that could get missed by skills mapping. 

3. Audit and map skills across your workforce.

If gauging employee confidence is the ‘why,’ then your AI skills audit should act as the ‘what.’ It’s less about understanding how your employees feel and more of a diagnostic that highlights the practical and operational realities of working with AI.

This means focusing on readiness, specific tool usage, and how well-informed and supported employees feel about using AI within organizational guidelines — and if they even know what those are.

Questions you can ask include:

• Which AI tools or platforms are you currently using as part of your everyday work?
• What types of tasks are you using AI for in your job? (e.g., drafting, data analysis, streamlining processes)
• Which of the following AI-related skills do you feel confident in? (List capabilities and skills as they apply to your organization.)
• Do you feel supported to use AI responsibly and in line with our organization’s policies?
• Do you know where to get support if you’re unsure how to use AI?
• What resources or support do you think would be most beneficial to you to improve your understanding of AI?
• Are there any AI-related skills you’d like to learn in the next 6-12 months?

Once you’ve compiled this data, it’s time to build an internal map of your organization’s skills. 

But the goal isn’t necessarily to create a detailed inventory. Instead, it’s more impactful to take an ontological approach and see how skills connect, overlap, and build the foundation for broader capabilities.

This helps you build skill clusters of related competencies that support specific types of work, like automation or predictive analytics. By mapping these clusters across teams, you don’t just see where the gaps are — you also see how employees can grow into adjacent skill sets and focus on targeted development efforts.

4. Track labor market data to identify new skills and emerging technologies.

Benchmarking skills internally is helpful, but you’ll also want to look for external signals to stay in step with how tech, jobs, and skills are shifting.

Global resources, like the World Economic Forum’s Future of Jobs Report 2025, provide a comprehensive overview of broader labor market skills trends. But if you want a more predictive signal, dig into broader labor market trends and run AI talent intelligence on competitors in your space. Consider the following questions:

• Which skills are competitors focusing on in job postings and employer branding?
• Which skills are slowing in demand? 
• How are technical and non-technical roles evolving?
• Which roles are being laid off?
• Where is compensation rising fastest and why?

Understanding how the market is shifting will help you spot emerging gaps and make a plan to develop the capabilities you don’t already have by promoting from within or hiring new talent.

Embedding AI Fluency into Talent and Learning Strategies

Creating long-term AI fluency takes more than a few upskilling initiatives or data analytics courses — it demands an organization-wide shift in processes, mindset, and behavior. 

Because what’s most important isn’t the tooling or training — it’s that you’re creating the environment for people to feel curious and safe using AI as part of their everyday work.

This relies on promoting experimentation, knowledge-sharing, and transparency about how AI will reshape careers and performance.

1. Create sandbox environments that promote safe experimentation.

Failure is part of learning how to use AI successfully. But often with AI, organizations equate failure with risk — meaning they impose tight controls on how and when employees interact with AI-powered tooling.

Organizations can maximize learning and creativity while minimizing risk by setting clear boundaries on what employees can and can’t do with AI and providing safe environments for experimentation.

“You don’t need to be a software engineer or understand machine learning to use AI,” Bradburn said. “But employees do need the opportunity to understand how the systems fit together. Go and experiment on genAI. Try building a website or creating an agent. Give them time to be curious, build their confidence, and learn through doing.”

Organizations can create the environment for AI-related experimentation by:

• Creating AI experiment logs and a build-in-public mentality so that employees can see how other teams are using AI, including successes and failures
• Cross-training teams from different functions to solve business problems with AI in hackathons
• Allocating protected time on the schedule for individuals and teams to do AI training programs or experiment with new technologies

2. Share use cases and knowledge across the organization.

One of the fastest ways to scale AI literacy is to show, not tell. When people can show how AI solutions apply to real-world contexts and demonstrate their business impact, AI adoption becomes tangible.

“Like with any other behavior within an organization where there’s significant change, you have to model it and show people what you’re doing,” Bradburn noted. “You have to surface examples of people on your team who are really pushing ahead with it, share how they did it, and the insights they got. AI isn’t unfolding evenly at all organizations — there are often huge gaps between managers and teams. But with strong knowledge-sharing, organizations can close these gaps.”

To scale the momentum for collective continuous learning, organizations can:

• Create community channels and message boards where teams can regularly trade generative AI prompts and use case tips, ask for guidance on AI projects, and share feedback in real time.
• Implement weekly office hours or an AI clinic where cross-functional experts across the company can help teams troubleshoot challenges.
• Create an accessible community library of approved use cases, prompts, and quick-reference guides for common AI tasks, tagged by department, model, or problem.
• Showcase high-impact use cases in company all-hands meetings.
• Create AI buddies within teams who can help mentor beginners around specific AI applications.

3. Create transparency around future growth and career paths.

One of the biggest unspoken tension points around upskilling your workforce with AI is the imagined face-off between humans and robots. 

Employees have very real fears that their livelihoods are under threat. This, said Fesinstine, is something HR and the C-suite need to address upfront — even if the reality is uncomfortable. 

“It’s critical to have real conversations with functional team members and employees with a sense of empathy, support, and real direction,” Fesinstine said. “You need to be able to say, this is what could be coming as we assess how technology and automation are going to change your role. Let’s create a strategy together so that we can identify where human impact is essential and support you for what’s next.”

Creating transparency isn’t about having all the answers. Instead, it’s about making sure you’re having open conversations with employees that help them visualize their future growth paths and next steps.

HR teams and managers need to be able to answer employee questions like:

• What are some potential future scenarios for my current role?
• What will the business need from my role or skills in the next few years?
• What skills or capabilities should I start developing now?
• What support is available to help me upskill or reskill?

{{rich-highlight-3}}

4. Update talent frameworks to reflect AI-driven behaviors.

As employees continue to experiment and grow with AI, organizations will also need to understand how AI adoption will reshape key employee lifecycle processes, like hiring criteria, performance reviews, and rewards strategies.

Take two employees in the same role as an example. Employee A has used AI to streamline customer queries and has seen a sharp rise in their customer satisfaction score. Meanwhile, employee B continues to answer customer queries manually, taking time and care to personalize their responses. Their satisfaction score remains high, but their overall case resolution speed is slower.

In a traditional system, employee A would likely be rewarded — with faster progression or a raise. So how can organizations best evaluate their progress fairly and equitably?
Fesinstine said this shift demands a new way to assess impact and potential.

“From a talent management perspective, we’ll need to start evaluating employee and candidate skills based on new criteria that reflect on how they’re using AI to generate either greater productivity or insights,” she said.

“It’s not just asking questions like ‘Can you do this or that skill or build a custom bot?’ It’s more like ‘How did this approach help you and allow you to expand X or Y?’ It’s not just recognizing AI tools — it’s how resourceful you are, and are you curious enough to explore?”

To make talent processes more equitable, organizations can:

• Define AI skills proficiency and fluency expectations by role or department.
• Track skills development, performance, and growth needs that contextualize individuals’ learning curves and effort.
• Help employees visualize career progression, skills development, and aligned competencies with platforms like Lattice Grow.
• Structure interviews based on AI potential, not fluency in specific skills.

Measuring AI Skills Gaps Over Time

When skills are constantly evolving, measuring what’s changing and where is like trying to gift wrap a cat — unpredictable and a little chaotic.

This is why structured, consistent processes that provide a holistic view of your organization’s progress with AI skills matter more than metrics:

• Integrate AI skills evaluation into performance processes. Use performance reviews, manager check-ins, and 360-degree feedback to monitor how employees are adopting and applying AI tooling and skills.

• Redefine performance and success metrics. Focus on outcomes, results, and learning behavior, rather than speed or volume.

• Track broader trends with analytics. Use an integrated platform like Lattice Analytics to centralize employee lifecycle data across processes and track broader trends in sentiment, hiring, performance, progression, and learning.

• Monitor learning goal progress. Align skills goals with business needs and track progress in dedicated dashboards to identify leading or lagging business segments.

• Create internal feedback loops. Use regular surveys, focus groups, and check-ins across teams to identify emerging skills gaps or friction points.

Where AI Skills Strategies Fall Flat

AI is still an evolving technology, which means organizations are going to make a lot of mistakes before they settle on the processes that work best. 

But there are some mistakes organizations can avoid off the bat to improve their success:

• Neglecting HR training: HR teams need the space and time to understand AI’s potential themselves before they can confidently roll out new processes and policies across the broader organization. Without this foundational knowledge, HR teams risk rolling out initiatives that date quickly — meaning employees look for workarounds that increase risk.

• Over-reliance on technical skills: Technical fluency is important, but organizations shouldn’t focus solely on this in isolation. Human skills like decision-making, critical thinking, evaluating risk, and using AI responsibly must be developed in parallel for long-term success.

• Measuring speed, not outcomes: With so many studies linking AI to enhanced workforce productivity, it’s tempting for organizations to measure its success in terms of time saved, cost savings, or volume. But what’s more impactful is how teams measure AI through an outcome-focused lens: results, capabilities or skills learned, team wellbeing, or improved decision-making. 

• Leaving managers unsupported: Managers are at the forefront of driving AI skills development and adoption within their teams. Providing targeted support through templates, coaching, and tooling will help them track skills penetration and monitor team sentiment at a team level.

{{rich-highlight-2}}