1. Trust and Explainability

    66% of employees criticize excessive trust in AI, and AI achieves only 6.2 out of 10 points on average.

    Trust Score (Future-Ready Workforce 2025)

    A lack of transparency becomes a credibility risk for managers and a barrier to employee acceptance.

    Solution: Model cards, explainability on demand, and clear traffic light rules (green = open, red = only under NDA)

2. Competence overkill vs. competence gap

    New hybrid roles (Prompt Engineer, AI Ethicist, Human-AI Interaction Designer) require a blend of expertise in technology, ethics, and

    communication.

    At the same time, there is a risk of being overwhelmed: Many employees feel over-promised and under-qualified by AI tools.

    Solution: Micro-learning paths, prompt coaching, and AI-saving sessions directly at the workplace.

3. Psychological safety and algorithm anxiety

   Employees fear de-skilling or surveillance through AI analytics

   82% of employees emphasize the importance of interpersonal relationships, but only 65% of managers share this

   priority – an empathy gap.

    Solution: Participative leadership, employee involvement in decision-making, AI retrospectives (moderated team meetings for

                     reflection on projects, with the goal of improving collaboration and work results), and clear no-go areas for

                     employee surveillance.

4. Role dissolution and loss of identity

   Instead of fixed job titles, there are dynamic role profiles that change every 6-12 months.

   The resulting tasks are modular, project-based, and platform-based – this disrupts traditional career patterns.

   Solution: Skill portfolios instead of job descriptions, internal gig platforms, and accompanying career coaching.

5. Ethical and liability pressures

    Executives are liable for discriminatory or unsafe AI outputs.

    External consultants deliver a black-box model – but internal managers bear the reputational risk.

    Solution: Mandatory second opinions (from experts and secondary AI systems), external fairness reviews, and contractual

                      explainability clauses.

6. Information Overload and „Continous Partial Attention“

    AI generates more data and options, not less work.

    Cognitive overload increases because humans have to iterate through prompts -> outputs -> prompts.

    Solution: AI filter boards, a maximum of one AI screen per meeting, and fixed deep work blocks without bot interruptions.

7. Uncertainty competence and leadership in permanent beta

    AI cannot predict the future – uncertainty remains.

    Leaders need the skills to navigate uncertainty: to make transparent decisions, even when data is incomplete.

    Solution: Scenario planning with AI, decision logs, and leadership training in ambiguity tolerance (the ability to accept, tolerate, and       

                     constructively deal with uncertainty, ambiguity, contradiction, or lack of clarity in various situations).

Short formula for decision-makers

Technology is only half the equation. The real challenges are cultural, psychological, and ethical—and they begin now, not in five years.

Addressing these seven areas transforms AI from a constraint to a co-intelligence partnership—leading to more resilient employees, more credible leadership, and future-proof organizations.

(Source: Adecco Group -Future Ready Workforce 2025; EY European AI Barometer 2025)

The discrepancy is not accidental – it is systemic and can be attributed to the following competing causes:

1.  Information Asymmetry

Executives derive their assessments of AI from strategy papers, consultant presentations, and pilot ROIs that promise efficiency and growth.

Employees experience AI as a tool update in their own workplace and primarily focus on fear of job loss, increased workload due to change, and a lack of training.

Therefore, 76% of German executives believe their teams are enthusiastic, while only 31% of employees confirm this – a ratio of almost 2.5:1 in favor of the cherry-picking perspective.

2. Different risk-affect logic

Group                       Main risk (top 1 concern)                                                  Justification

________________________________________________________________________________________________________________________________________________

Leaders                   Face a competitive disadvantage if we don't scale AI     They see AI as a strategic lever – the biggest risk is: 

                                                                                                                                    being forced out of the market faster than by employees.

Employees              My job disappears or becomes devalued                         They experience AI not as an advantage, but as an immediate threat

                                                                                                                                    to their work, their salary, and their identity.

The specific dominant fear determines which information is perceived, which stories are believed, and which measures are demanded -therefore, it is the main lever for any change strategy.

This differing level of concern explains why 77% of employees specifically fear job loss, while less than a third of executives believe their employees even share this concern.

3. Communication filters and change fatigue

Top-down slides promising AI benefits often fail to convince employees because they coincide with restructuring, salary freezes, or layoffs.

A lack of participation (we were never asked) turns skepticism into outright rejection—even leading to shadow IT or active boycotts of AI tools.

4. Different skill realities

94% of employees say: I am ready and motivated to learn about AI.

Only 5% of companies offer systematic training. Managers are not taking enough time, or any at all, to train and equip employees for working with AI. The consequence: employees are frustrated and demotivated.

The gap between willingness to learn and available training opportunities exacerbates feelings of isolation – and thus increases the anxiety statistics (data privacy concerns 63%, burnout concerns 60%).

Conclusion: Two films on one screen

The discrepancy arises not primarily from ill will on the part of management, but from:

1. Lack of feedback loops (real-world insights from the ground up)

2. Different incentive systems (shareholder value vs. job security)

3. Lack of capacity in change communication and learning infrastructure

4. Shortage of experienced managers because experienced people were laid off during restructuring

Those who fail to address these four key factors overestimate enthusiasm and underestimate resilience – with the result that transformation projects stall.