Shadow AI and Training Gaps: What the Latest SAP Study Means for SMEs

A new SAP-sponsored study conducted by WalkMe has raised important questions about how AI tools are being used inside organizations - and more importantly, how well people are trained to use them.

New SAP study by WalkMe uncovered some interesting topics concerning how AI tools are being used inside companies/organizations by employees.  

Here are some of the key findings:

• 73% of employees are using AI tools without the knowledge or approval of IT.

• Only 7% of employees say they are fully trained on the digital systems they use every day.

• 49% of respondents say their organizations don’t track how software is being used post-deployment.

• Many organizations struggle to measure the ROI of digital transformation initiatives.

In this blog, we’ll explain what the findings mean, especially for small and mid-sized enterprises (SMEs) and how you can respond to it.

‍What Is Shadow AI?

First things first - Shadow AI refers to employees using artificial intelligence tools outside of official channels. Most common examples are:  

• Using ChatGPT or similar tools to generate reports or emails.

• Automating spreadsheets using scripts or third-party plug-ins.

• Using personal AI tools with company data without approval.

There was a similar term before AI - “shadow IT,” where employees use unapproved hardware or software in the workplace. Now, however, the risks are even higher due to AI’s speed and influence.

‍

Why Is Shadow AI Growing So Quickly?

There are several reasons why this is becoming a problem so fast:

1. AI tools are easy to access - Anyone with a browser or a smartphone can use ChatGPT, Gemini, or other AI tools.

2. Employees want to be more efficient or avoid repetitive work – Employers often need to complete tasks on short deadlines and AI can be helpful with that. Going through large amounts of data can be much easier with AI summarizing the data for you.  

3. Official systems may lag behind – Let's be honest, enterprise systems are not often the most flexible and fast. This may push some to look elsewhere for simple tasks that take a bit too long in the official and often outdated systems.  

The whole thing is quite natural to humans – we like shortcuts. And the intention is almost always good – we’re trying to save time or improve our workflows. However, it is also true that without structure, the benefits can be limited and the risks real.

‍

Why Shadow AI Is a Problem for SMEs

For smaller businesses, this issue carries some risks:

• Inconsistent processes: When different employees use different tools without coordination or any kind of structure, it leads to fragmented workflows.

• Data security concerns: Sensitive data might be entered into tools that aren’t compliant or secure.

• Reduced accountability: If something goes wrong, it’s harder to track where or why.

• No visibility for leadership: If you can’t see the problem, you can’t improve it.

In other words, you might think your organization is “using AI” — but without a plan, you might not be getting any measurable value from it.

‍

The Bigger Picture: Training Gaps in Digital Tools

The survey also found that the lack training is a major issue in how digital tools are adopted.

Only 7% of employees reported feeling fully trained on the tools they use. And in many cases, training only happens during implementation.

This creates a few challenges:

• New hires may never receive formal onboarding.

• Teams don't know how to use newer features as systems evolve.

• Software adoption stalls, therefore ROI suffers.

For SMEs, where budgets are tight and staff are often multitasking, training tends to be informal if any. And what is the point of having a tool if you don’t know how to properly use it?

‍

What This Means for Digital Transformation

Digital transformation depends on three pillars:

1. Tools (the platforms and software you use)

2. People (your teams who use them)

3. Processes (how tasks are defined and executed)

If you’re investing in tools but not empowering people or structuring processes, you won’t get the ROI you expect.

The SAP-WalkMe study highlights this gap clearly: tech adoption alone isn’t enough. Businesses must also invest in usability, visibility, and learning.

‍

How SMEs Can Respond: 5 Practical Steps

It doesn’t really matter if you have officially started a process of AI adoption, chances are that your employers are using it in one form or another. So, technically you are already adopting AI . So Here are five steps SMEs can take based on the study’s findings:

1. Audit Current AI and Automation Use

• Identify tools being used across departments.

• Look for AI usage that may be happening unofficially.

• Check for redundancies or potential risks.

This will help you to understand where you are and the scope of “shadow AI” in your organization.

2. Centralize Tooling Where Possible

• Adopt platforms that integrate with your existing systems.

• Avoid using too many “single-purpose” AI tools.

• Establish IT-approved tools that offer flexibility but remain secure.

These are the first steps to help you eliminate fragmentation and encourage shared usage patterns.

3. Assign Ownership for AI and Automation Strategy

This could be someone in operations, IT, or finance - but someone needs to own and be the point person for AI usage to monitor outcomes and manage risks.

Even in small teams, a part-time role dedicated to automation/AI can go a long way.

4. Implement Training as an Ongoing Practice

• Offer short, role-specific training sessions on key tools.

• Don’t assume that one-time onboarding is enough.

Since AI tools are evolving quickly you need to ensure that your team’s knowledge does as well.

5. Track Outcomes and Adjust

• Use KPIs to measure success: time saved, errors reduced, accuracy improved.

• Collect feedback from users to improve workflows.

• Adjust and improve based on feedback.  

Automation and AI is iterative and it is always ongoing.  

‍

Final Thoughts: Structure Beats Speed

The survey doesn’t suggest that AI is a bad idea - on the contrary. It just shows that when AI is adopted without structure, businesses miss out on its full potential.

SMEs should take this opportunity to review their current digital landscape. Ask:

• Are we using tools effectively?

• Do our people feel confident using them?

• Are we seeing measurable improvements?

If the answer is unclear, it’s time to take a more structured approach.

‍