netzstrategen AI Operations.
Insights Glossary Resources Deutsch Self-Check
← All posts
Enablement

Fascination Meets Fear of Losing Control: What AI Skepticism Means for Companies

Published on 8/31/2026 · André Hellmann

Allensbach Adoption Enablement Change

AI acceptance in the workforce decides the outcome of every AI rollout. Employees are not a special group. They are the general public — and the public holds two views of AI at the same time. A survey by the Allensbach Institute for Public Opinion Research, commissioned by the FAZ, makes that split measurable (Source: IfD Allensbach for FAZ, fieldwork 5–17 June 2026, n=1,047).

Where do you stand?

Discuss your next step in a free diagnosis call. Book a slot →

Contents

Two feelings, one technology

Most people in Germany consider AI useful. And unsettling. Both at once.

77 percent associate AI with “progress”, 70 percent with “helpful”, 55 percent with “effectiveness”. At the same time, 55 percent fear a loss of control, 53 percent call AI “opaque” and 44 percent “eerie” (Source: IfD Allensbach for FAZ, 2026).

That is not a contradiction. It is a judgment without orientation. People see the benefit and recognize that they do not understand what happens inside the system.

For companies, this is the starting point of every AI rollout. Not rejection. Not enthusiasm. An open question that needs an answer.

What the Allensbach survey shows

Allensbach surveyed 1,047 people between 5 and 17 June 2026 on behalf of the FAZ. Two findings matter for companies.

First: usage is already high — and extremely uneven. 41 percent of the population use AI programs frequently, 30 percent rarely, 29 percent never. Among 16- to 29-year-olds, 69 percent use AI frequently and only 6 percent never. Among people over sixty, 52 percent never use it (Source: IfD Allensbach for FAZ, 2026).

Second: the skepticism targets transparency, not performance. “Opaque” and “loss of control” say nothing about output quality. They describe a missing explanation.

One team, three AI realities Use of AI programs by age group, share in percent Age 16 to 29 69% frequently Total population 41% frequently 30% rarely 29% never Over 60 years 52% never frequently rarely never Youngest and oldest group live worlds apart — inside the same company. Chart: netzstrategen · Source: IfD Allensbach / FAZ, 2026 (n=1,047) netzstrategen
Use of AI programs by age group. For the over-sixty group the survey reports the "never" share; the remaining shares are not published.
For presentations:

A third finding is worth noting. 71 percent reject the idea that broad general knowledge has become unnecessary thanks to AI. Only 19 percent agree — though among younger respondents the figure rises to 37 percent (Source: IfD Allensbach for FAZ, 2026). People want to use AI without handing over their own judgment.

A free Self-Check shows where a company stands on acceptance and enablement in a few minutes.

Why a split public view lands inside the company

Any AI rollout happens inside exactly this population. These numbers are not a societal mood chart. They describe the workforce.

Three operational consequences follow.

  • A rollout hits three groups at once. Some have used AI privately for years. Others have never opened a tool. A single training format serves neither well.
  • Skepticism rarely shows up as objection. It shows up as silent non-use. Access is granted, the license runs, the tool stays closed.
  • Introducing AI without explanation confirms the fear. “Opaque” is not a prejudice when nobody explains which data goes where and who decides in case of doubt.

Acceptance does not come from persuasion. It comes from transparency.

The pattern is familiar. It is the people-process gap: the technology is in place, daily work stays the same. The Allensbach numbers now supply the population-level evidence.

There is an aggravating factor. Where AI is in place, 47 percent of employees spend more time managing AI than doing their actual work (Source: BCG, AI at Work, 2026). Anyone who was skeptical sees that confirmed. Why this happens is covered in the article on the tool-stacking trap.

From skepticism to confidence: enablement as an operating discipline

The obvious answer to skepticism is communication. It falls short. People who fear losing control need control — not a campaign.

So enablement in our work means one thing: people get tools, boundaries and responsibility at the same time. Four building blocks belong to it.

  1. Transparency about data. Which data leaves the company, which does not? That answer comes first, not in an appendix.
  2. Clear decision rights. Which outputs go out unchecked, which do not? Whoever approves AI suggestions keeps control — and feels it.
  3. Practice on the real workflow. People train on their own task, not on a slide-deck example.
  4. One visible benefit per person. A task that is demonstrably faster. Without that proof, training stays theory.

The difference between explaining and enabling is measurable. A clear AI strategy lifts impact by 25 percentage points, better tools alone by only 5 percentage points (Source: BCG, AI at Work, 2026). Strategy and enablement beat tooling.

For leadership this creates a dual task: rebuild the operation and bring people along. Both at once. The article on the dual leadership challenge in AI transformation sets out what that involves.

Built with the team: how to start

We build AI systems with the team, not for the team. That is not a matter of attitude but of acceptance. People who helped build a system know its limits — and do not distrust it.

In practice this means three things:

  • The workflow comes from the people who run it daily. They know where time is lost. No audit replaces that.
  • Cockpits instead of open-ended prompting. A clear interface with defined tasks takes the mystery out of using AI.
  • Enablement as a fixed step, not an afterthought. Training and enablement are step 5 of our engagement steps — before go-live, not after it.

How that approach works in detail is described in Built with the Team.

Which workflow suits a first step is something we map out in the free diagnosis call.

Conclusion: acceptance is an operating metric

The Allensbach numbers do not describe hostility to technology. They describe a population that uses AI and does not see through it.

The uncomfortable consequence for companies: acceptance cannot be communicated into existence. It emerges where people understand what a system does and get to decide what happens with its output.

That is work on the operation, not on the image. Transparency about data, clear decision rights, practice on the real workflow, one proven benefit per person. Four building blocks that belong before the rollout — not after it.

Frequently asked questions about AI acceptance

How do people in Germany view AI?

Largely positive, with clear reservations. 77 percent associate AI with “progress”, 70 percent with “helpful”. At the same time, 55 percent fear a loss of control and 53 percent find AI “opaque” (Source: IfD Allensbach for FAZ, 2026).

How many people use AI programs regularly?

41 percent of the population use AI frequently, 30 percent rarely, 29 percent never. Among 16- to 29-year-olds, 69 percent use it frequently; among people over sixty, 52 percent never do (Source: IfD Allensbach for FAZ, 2026).

Why do AI rollouts fail inside the workforce?

Not through rejection, but through silent non-use. Access is granted, daily work stays unchanged. That is the people-process gap: technology without a changed workflow produces no value.

What helps against skepticism in a team?

Control rather than communication. Transparency about data flows, clear decision rights, practice on the person’s own task and one proven benefit each. A clear AI strategy lifts impact by 25 percentage points, better tools alone by only 5 percentage points (Source: BCG, AI at Work, 2026).

Where should a company start with enablement?

With a workflow the team itself names as a time sink. Which one that is in a specific case is something we clarify in the free diagnosis call.

Sources

What's next