Out of the Shadows, Into Control: Why Companies Need AI Operations Now

Shadow AI describes AI tools that employees use at work without the company having provided them or knowing about them. They usually run on personal accounts and free versions. Shadow AI rarely comes from bad intent; more often there is simply no tool supplied by the company. New ZEW data shows how widespread this situation is: almost no company bans AI, but many allow it without providing anything (Source: ZEW, Information Economy Sector Report, 2026).

4%

fully prohibit AI (information economy)

ZEW, 2026

16%

explicitly permit AI but provide nothing

ZEW, 2026

23%

same case in manufacturing

ZEW, 2026

Contents

• What shadow AI is and why it happens
• Where shadow AI starts: allowed but not provided
• Step 1: Make shadow AI visible
• Step 2: Resolve shadow AI
• Why this is enablement, not IT
• Conclusion: out of the shadows
• Sources

What shadow AI is and why it happens

Shadow AI means AI being used in a work context without any company-managed access behind it. Typical examples are personal accounts with chat assistants, free versions of writing and translation tools, or browser extensions that nobody has approved.

The reasons are usually practical. People want to get a task done faster, have had good experiences in private, or see colleagues doing the same. Without a company tool, they reach for their own.

The problem is less the use itself than its invisibility. Whatever runs in the shadows, a company can neither secure, nor evaluate, nor improve.

Where shadow AI starts: allowed but not provided

The debate about shadow AI usually revolves around bans. Yet bans are rare. In March and April 2026, ZEW surveyed around 1,500 companies from the information economy and manufacturing. Full bans are the exception: 4 percent in the information economy, 8 percent in manufacturing (Source: ZEW, 2026).

It gets revealing once permission and provision are examined separately. ZEW does just that and finds four states.

For presentations:SVGPNG

In the information economy, 58 percent of companies actively provide generative AI to their staff. In manufacturing it is 30 percent (Source: ZEW, 2026).

The remainder splits across three states, two of which are problematic. 16 percent in the information economy and 23 percent in manufacturing explicitly permit use without providing an application. A further 22 and 39 percent respectively tolerate use without clear guidance (Source: ZEW, 2026).

Added together, this means that in manufacturing, 62 percent of companies have AI in use, or permitted, without supplying a tool for it. This grey zone is where shadow AI grows.

Permission without a tool is not governance. It delegates the risk to the workforce.

Step 1: Make shadow AI visible

Anyone who wants to resolve shadow AI first needs to know what is happening in their own company. “Allowed but not provided” leaves employees alone with the task. That has three predictable consequences.

• Personal accounts take over. Someone expected to work faster, with no company tool available, uses their own. At that moment the task leaves the company’s contractual perimeter.
• Copy-paste becomes a data path. Tender documents, quotation calculations, customer correspondence and design specifications travel through the clipboard into systems whose processing terms nobody has reviewed.
• Nobody knows what is running. Without provided access there is no usage data. That removes the basis for any evaluation, including the question of where AI already creates value in the company today.

So making it visible starts with asking. A short, anonymous survey per team is enough to begin with: which tools are in use, for which tasks, and with which data? The answers show where the need is greatest and which workflows need provided access first. Threatening sanctions along the way gets you no honest answers.

Step 2: Resolve shadow AI

Data protection is a good reason to resolve shadow AI. Data protection concerns are among the most common reasons for not using AI at all: 60 percent of companies without AI cite them (Source: Destatis, 2025). Yet at other companies the same concern leads to no provision at all instead of secure provision. The result is higher risk, not lower.

The more important reason lies elsewhere. Companies that actively integrate AI into their workflows move topics forward deliberately and make full use of their people’s potential. Shadow AI shows that this potential exists. What is missing is a framework in which it can take effect.

That does not take a twelve-page policy. Four decisions are enough to start, and they can be made in days.

1. A named account. One company account per person, billed through the company’s contract. This is the only item that costs money, and the only one that actually closes the grey zone.
2. Data classes instead of prohibition lists. Which data may enter an AI system, which may not? Three classes suffice: non-critical, internal, confidential. Without this split, every individual decides alone.
3. A short list of approved tools. Two or three applications, not twenty. Short lists get read, long ones get ignored.
4. A named contact person. Someone to ask before uploading something. Without that address, people do not ask. They act.

These four cost less effort than the debate about whether to permit AI at all. How to set up data flows cleanly is described in the article on AI and data protection.

Turning provision into lasting operation takes more than an account: clear responsibilities, approvals, usage data and regular review. We call this framework AI Operations.

Why this is enablement, not IT

The gap is often handed to IT. That falls short, because IT answers the wrong question.

IT can create accounts, review contracts and secure data flows. What it cannot do is decide which workflow in sales or engineering actually needs AI support. That decision belongs in the business unit.

So provision is an enablement task. It covers three things beyond technical access: a concrete use case per team, practice on a real transaction, and named accountability for the results.

That is how we build: with the team, not for the team. Why this is a question of acceptance rather than attitude is described in Built with the Team. How decision rights can be graduated is covered in Governance, Control, Autonomy.

Which workflows need provided access first is something we map out in the free diagnosis call.

Conclusion: out of the shadows

The ZEW figures clear up a misconception. Companies barely ban AI. They permit it and supply nothing.

That is the worst of the three possible states. A ban is at least a decision people can align to. Provision is a decision with a tool attached. Permission without a tool is neither: it pushes the decision about data paths onto every individual in the company.

The way out has two steps. First, make visible what is running. Then resolve it with one account, three data classes, a short tool list and one name. With that in place, use comes out of the shadows and AI can be deployed where it delivers the most value.

This article gives a general overview and does not replace legal advice. For data protection and employment law questions, contact your data protection officer or legal counsel.

Sources

• ZEW, 2026: Information Economy Sector Report, May 2026 (survey March/April 2026, around 1,500 companies)
• ZEW, 2026: Hardly Any Companies Forbid the Use of AI, 8 June 2026
• Federal Statistical Office (Destatis), 2025: Use of ICT in enterprises: reasons for not using AI, 2025 survey