The AI Implementation Gap is the real problem for small and medium-sized enterprises (SMEs). A recent study sums it up: AI is ready, but companies are not. Those who still treat AI purely as a technical issue are fighting on the wrong front. The tools work. What is missing is organizational integration. This is precisely what now decides competitiveness, customer loyalty, and even whether employees stay.
What the Numbers Show
In June 2026, Thomson Reuters published its fourth "Future of Professionals" report, based on a survey of 1,816 professionals in law, tax, audit, compliance, and risk in 62 countries.1 The study primarily concerns professional service providers, not the classic industrial SME sector. However, the pattern can be almost directly transferred.
Key findings of the report:
- Usage is not the problem. 74 percent of respondents already use AI tools weekly. Nevertheless, 91 percent say their organization is not fully leveraging the potential.1
- Strategy and implementation diverge. Even where an AI strategy exists, 35 percent say that ambitions are not reflected in daily work. Almost one in five lacks a clear strategy altogether.1
- The gap costs talent. One in four professionals (24 percent) who experience a disconnect between what is technically possible and what is delivered within the company considers quitting within two years. Almost half of executives, however, believe that pressure on the labor market is still at least three years away.1
- The gap costs customers. 78 percent of business customers consider AI-supported quality improvements to be very important or indispensable. Only 6 percent see them delivered by their service providers. Almost a third plan to review provider relationships in the next twelve months. In the US legal and accounting market alone, this puts approximately 143 billion dollars at stake.1
A third point leads directly into the governance area: a third of professionals use AI tools that their organization has not approved, known as Shadow AI. In companies considered too slow, this figure rises to 41 percent.1 Secret tool usage is not a discipline problem. It is a symptom of a gap between what employees need and what the organization provides.
Why This Also Affects German SMEs
One might argue that this is an Anglo-Saxon service provider issue. A look at German figures shows the opposite. According to the ifo Institute, as of May 2026, 54.5 percent of German companies already used AI, up from 40.9 percent the previous year.2 AI has "finally arrived in the mainstream," according to the institute.
The distribution is crucial: large companies are at 67.2 percent, small ones at 51.2 percent, and SMEs are furthest behind at 47.2 percent.2 Precisely SMEs, which should thrive on lean implementation, are lagging. The real question today is: Are we actually gaining value from this usage? This is precisely the AI implementation gap.
The Gap Is Organizational, Not Technical
If 91 percent are not exploiting the potential, even though three-quarters use the tools weekly, it's not due to a lack of tools. It's due to what stands between the tool and value creation:
- No one has clarified responsibilities: Who decides on use cases, who is liable for errors, who provides training?
- There are no approved, auditable tools, so employees resort to shadow tools and transfer confidential data externally. Our article on Governance and Law in AI shows how to regulate this properly.
- A framework is missing to turn isolated usage into repeatable processes. Individual employees who cleverly operate an AI tool do not yet make a productive organization. The article Too Many AI Tools, No Priority helps with prioritization.
- Iteration is missing: Pilots are launched but not measured, not improved, not scaled. Thus, projects get stuck in the familiar pilot trap.
These are not technical questions. They are questions of strategy, governance, organization, and process. This is exactly why we built the 6 Rocks: a framework that covers all dimensions of an AI transformation.
What You Can Do Specifically
You don't have to close the gap in one large project. But you should start before customers and employees draw the consequences for you. Four steps that create value in any order:
- Create an AI inventory. Get an overview: Which tools are in use, officially and unofficially? Where is Shadow AI running? This is the basis for everything else and can be done in a few days.
- Clarify responsibilities. Appoint a person responsible for AI within the company. This person is the contact for prioritization, approvals, and training.
- Provide approved tools. Offering auditable, data protection-compliant tools dries out Shadow AI. This lowers the risk more effectively than any prohibition. Prohibitions only drive usage underground.
- Execute one use case thoroughly. Select an application, measure its benefit, improve it, and only then scale it. You can read how a phased approach looks in Introducing AI in Phases.
The common thread remains: the implementation gap closes through structure. And structure doesn't create itself. It must be built and then maintained.
Conclusion
The study data is clear: AI has arrived in SMEs, but its value contribution has not. Those who now clarify responsibilities, approve tools, and thoroughly complete use cases will gain a lead that is hard to catch up with, while others are still discussing tool selection. The gap is real, but bridgeable. The first step is a clear framework.
Do you want to know where the biggest implementation gap lies in your company? In a 30-minute AI check, we assess your status along the 6 Rocks and identify the three levers with the quickest value contribution. No sales pitch, no slides. A structured discussion for management and AI stakeholders.
Sources & References
- Thomson Reuters, "AI is Ready but Firms are Not", Press release for the Future of Professionals Report 2026, 22.06.2026: thomsonreuters.com
- ifo-Institut, AI use by German companies (May 2026): ifo.de