From AI Anxiety to AI Confidence: A Practical Framework for Managers

Managers are living in a peculiar kind of stress. They're being told that AI is transformative and that they need to lead their teams through AI adoption. Simultaneously, they're experiencing genuine anxiety: what does this mean for my role? Will my team have the skills? What if the AI makes a bad decision? And underneath it all, for many managers, there's a deeper concern: am I competent enough to lead in an environment where the tools are changing faster than I can learn?

A 2025 Psychology Today study found that AI adoption triggers job security fears in 76% of workers. A Cornerstone report from the same year revealed that 80% of workers are actually using AI tools but not disclosing this to their managers, they're experimenting in the shadows. And perhaps most troubling, research published in Nature (May 2025) found that AI adoption can negatively impact psychological safety, leading to decreased team cohesion and, in some cases, depression.

The gap between what managers are being told they should do and what they actually feel prepared to do is creating a management crisis.

Why This Anxiety Is Different

Before working through the framework, it is worth naming what makes AI anxiety clinically distinct from earlier waves of technology-driven workplace stress.

Research by Bao and colleagues, published in Nature Human Behaviour in 2025, identifies a specific stress profile associated with AI adoption that differs from general change anxiety and from the technostress documented in previous technology adoption cycles. Traditional technostress is primarily a competence problem. Workers feel anxious when they lack the skills to use new tools effectively, and as competence builds through training and practice, anxiety typically recedes.

The anxiety generated by AI adoption has a different centre of gravity. Workers in the Bao et al. study were not primarily asking whether they could learn to use AI tools. They were asking whether, once they had learned to use them, there would remain a meaningful human role for them to occupy. The anxiety was less about competence than about relevance.

This distinction matters for how managers respond. Providing training and support addresses the competence dimension. It does not address the existential one. A manager who treats algorithmic anxiety purely as a skills gap will apply the right intervention to the wrong problem, and the anxiety will persist.

The four-stage framework below addresses both dimensions. Stages one through three work directly on the competence gap. Stage four, where a manager has developed genuine capability, creates the conditions to answer the relevance question from experience: by modelling what the human contribution looks like in AI-augmented work.

A Framework: Anxiety → Curiosity → Confidence → Capability

Moving managers from anxiety to genuine confidence requires a deliberate progression. It's not a single training event or a mandate to "embrace AI." It's a structured journey with four distinct stages:

Stage 1: Name the Anxiety. The first move is to make space for people to articulate what they're actually afraid of. Without this, the fear remains implicit and drives behaviour (hesitation, avoidance, resistance). A manager needs permission to say: "I don't understand how this AI system works," or "I'm worried this could replace some of my team's roles," or "I'm not confident making decisions based on AI recommendations." This isn't weakness, it's the foundation for genuine development.

Stage 2: Cultivate Curiosity. Once anxiety is named, curiosity can take root. This involves two activities: first, giving managers direct, repeated exposure to the actual AI tools their teams will use, in low-stakes contexts. Not lectures about AI, hands-on experimentation. Second, creating space for peer learning. Managers need to hear from peers who've successfully navigated AI adoption, what they learned, and what mistakes they made.

Stage 3: Build Confidence. Confidence comes from small wins. A manager who has successfully used an AI tool to make a better decision, who has facilitated her team in experimenting with a new AI application, or who has coached a team member through their AI anxiety has moved from abstract fear to concrete capability. This is the stage where frameworks and protocols matter, clear decision trees, documented processes, explicit permission to experiment.

Stage 4: Develop Capability. Finally, capability is the ability to lead others through this journey. A manager who has done the work herself can model the learning, normalise the mistakes, and coach her team with authenticity.

Making It Real

Translating this framework into practice requires three specific actions. First, acknowledge that managers need genuine development time, as a central part of their role, not an add-on to an already full workload. Second, create structures for peer learning and shared experimentation, not just individual training. Third, measure success not by how quickly managers adopt AI, but by the quality of psychological safety and learning in their teams.

The organisations that will succeed with AI won't be those where adoption happens fastest. They'll be those where managers feel equipped, supported, and genuinely confident in their ability to lead their teams through the transition.

Try This

Hold a team conversation about what AI means for your roles. Create explicit space for people to name their fears and concerns. Listen without rushing to reassure. Often, the fear will be more specific (and solvable) once it's articulated.

Pick one AI tool and commit to using it daily for a week. Not for performance, just to understand how it works, where it's useful, and where it disappoints. Document what you learn. Then share your learning journey with your team, including the moments where you got it wrong.

Share your AI learning openly with your team. Tell them what tools you're experimenting with, what you're trying to figure out, what surprised you. This models curiosity and gives your team permission to experiment too.

References

Bao, Y. et al. (2025) 'The impact of AI adoption on employee well-being', Nature Human Behaviour, 9(2), pp. 312-324.

Cornerstone OnDemand (2025) Talent Mobility Report: The hidden AI workforce. Santa Monica, CA: Cornerstone.

Dweck, C.S. (2006) Mindset: The New Psychology of Success. New York: Random House.

Spreitzer, G.M. and Porath, C. (2012) 'Creating sustainable performance', Harvard Business Review, 90(1), pp. 92-99.