From Threat to Creative Fuel: How Teams Can Harness AI Anxiety
Part Three of a Five-Part Series on AI Anxiety in the Workplace
In my last article, I explored how AI anxiety spreads through teams through a form of social contagion that can quietly turn a few people’s worries into a collective drag on wellbeing and performance. Understanding that contagion matters. But it also raises a more practical question: what are we supposed to do about it?
The default response in many organisations is to try to reduce anxiety - calm people down, reassure them, encourage them to relax. The difficulty is that this approach rarely works, and when it does, it can backfire. Telling someone who is anxious to “just calm down” often makes them feel worse. Reassuring people that “your job is safe” sounds hollow when no one can honestly guarantee that. And even if we could remove anxiety entirely, we might not want to. The energy behind that anxiety can be highly productive if we learn how to channel it.
So this article takes a different angle. Instead of trying to suppress AI anxiety, we can help teams transform it, using the physiological arousal that comes with anxiety as fuel for creativity, learning and better performance. The crucial insight is that anxiety and excitement are not opposites. Neurologically, they are remarkably similar states, interpreted in different ways (Brooks, 2014).
The Arousal Paradox: Anxiety and Excitement Share the Same Wiring
Ask people what they should do when they feel anxious before an important presentation and most will say, “try to calm down”. This advice is so culturally embedded that “Keep Calm and Carry On” became one of the most recognisable slogans of the twentieth century. The problem is that calming down on demand is extremely difficult — and research suggests it may be the wrong goal anyway.
Alison Wood Brooks at Harvard Business School tested this idea in a series of experiments (Brooks, 2014). Participants completed anxiety-inducing tasks such as singing karaoke, giving impromptu speeches and solving maths problems under time pressure. Some were told to say “I am calm” beforehand; others said “I am excited”. Those who reframed their anxiety as excitement consistently performed better. They sang more accurately, spoke more persuasively and solved more problems correctly.
Why does this work? Psychologists refer to arousal congruency. Anxiety and excitement are both high-arousal states: heart rate rises, stress hormones increase and attention sharpens. When you are already physiologically activated, trying to force yourself into calm — a low-arousal state — requires a major biological gear change. Shifting from anxiety to excitement, however, requires very little physiological change. The body is already energised; what shifts is the interpretation (Brooks, 2014).
More recent meta-analyses reinforce this. Interventions that help people reinterpret stress responses as helpful rather than harmful are associated with improved performance outcomes (Bosshard & Gomez, 2024; Hase et al., 2025). In other words, how we label our arousal shapes what we can do with it.
Challenge Versus Threat: The Biopsychosocial Model
To understand why some people perform well under pressure while others struggle, researchers developed the biopsychosocial model of challenge and threat. The model suggests that in demanding situations we rapidly (and mostly unconsciously) assess both the demands we face and the resources we have to meet them.
When we believe our resources are sufficient, we enter a challenge state. The cardiovascular system responds efficiently: the heart pumps more blood and blood vessels dilate, supporting approach behaviour. We lean in, stay flexible and see more options.
When we believe demands exceed our resources, we enter a threat state. Blood vessels constrict and the body prepares for defensive action. Attention narrows, thinking becomes more rigid and we focus on avoiding losses rather than exploring possibilities (Seery, 2011).
What is striking is that the external situation does not determine which state we enter. Two people in the same meeting about AI adoption can have entirely different physiological and psychological responses depending on how they appraise the situation.
For teams facing AI change, this is a crucial insight. The technology itself does not automatically create threat or challenge. What matters is whether people feel they have the resources, skills, time, support and psychological safety, to cope. Leaders who only try to reduce demands (by slowing rollouts, for example) are addressing only half the equation. Building perceived and actual resources is just as important.
From Individual to Collective: Shifting Team States
Most challenge–threat research focuses on individuals. Teams, however, operate through shared emotional climates. If anxiety can spread socially, more constructive states can too.
Research on team psychological safety shows that when people feel safe to take interpersonal risks, they are more likely to speak up, experiment and engage in creative problem-solving (Jin & Peng, 2024). In one recent study of employees in high-tech organisations, team psychological safety was strongly linked to innovative performance, with communication playing a key mediating role (Jin & Peng, 2024). When people can openly say, “I tried this and it didn’t work,” or “I’m excited about what this tool might do,” collective learning accelerates.
Psychological safety alone, though, is not enough. Teams also need a shared challenge framing — a collective story that AI adoption is a stretch and a learning journey, not simply a threat to survival. This does not mean denying risks. It means deliberately shaping the meaning of the moment.
Four Practices for Turning Anxiety into Creative Energy
Based on the research and my work with teams internationally, four practices consistently help teams harness rather than suppress AI anxiety.
1. Name and Normalise the Arousal
AI genuinely triggers physiological arousal. Pretending otherwise makes people feel isolated. Overstating the danger amplifies fear. The middle ground is honest normalisation: this is significant, it is understandable to feel energised or unsettled, and that energy can be useful.
When people recognise their arousal as appropriate rather than a personal failing, the secondary anxiety - worrying about being anxious - reduces. The energy remains, but it feels less threatening.
- Build Collective Experimentation Rituals
Perceived resources grow through experience. One of the most powerful resources a team can build is the ability to experiment together.
When experimentation is regular, low-stakes and shared, several shifts occur. The novelty of AI fades. Competence builds through small wins. Struggles become visible and normal rather than private and identity-threatening. And quite often, curiosity and excitement begin spreading just as quickly as anxiety once did.
In the teams I work with, the turning point often comes when exploration becomes a routine team activity rather than something individuals feel they must master alone.
3. Celebrate What AI Cannot Do
AI anxiety is often, at heart, identity anxiety. If the machine can do parts of my job, what is left that is uniquely mine?
One counterbalance is to deliberately recognise and develop distinctly human capabilities: relationship-building, ethical judgement, contextual awareness, creative leaps across domains and emotional attunement. This is not about denying AI’s strengths. It is about keeping a balanced view of both human and technological contribution.
Practically, this might mean recognising moments where human judgement made a difference, discussing how roles are evolving, and investing in the capabilities that remain deeply human.
- Reframe the Learning Curve
In many workplaces, not knowing how to use AI yet feels like a deficiency. That interpretation almost guarantees a threat response. An alternative is to treat the learning curve as the point. “We’re still figuring this out” can signal possibility rather than inadequacy. This connects with growth mindset research but extends it to the team level: the group develops a shared story about learning, experimentation and adaptation.
When the narrative shifts from “we should already know this” to “we are building this together”, the emotional tone changes with it.
When Reappraisal Is Not Enough
Reframing anxiety as excitement does not change material realities. If jobs are genuinely at risk, if change is being handled poorly, or if trust in leadership is low, cognitive reframing alone will not solve those problems.
What reappraisal can do is help people function more effectively in situations that are uncertain but not necessarily catastrophic. It creates the emotional and cognitive bandwidth needed to engage, learn and adapt. Without that shift, anxiety about AI can make it harder to learn about AI, which in turn fuels more anxiety.
Leaders who use these practices must also do the harder structural work: creating real psychological safety, aligning AI adoption with human wellbeing and building trust. Without those foundations, reframing risks feeling manipulative rather than supportive.
The Creative Potential of Uncertain Times
The heightened arousal many teams feel around AI is not simply a problem to eliminate. It is energy. The same activation that underlies anxiety can also support focus, creativity and growth if the context supports a challenge rather than a threat response.
The teams that thrive will not be those with the least anxiety. They will be those who learn to use that collective energy as fuel for experimentation, learning and imaginative engagement with new tools. The shift from “How do we calm people down?” to “How do we help people use this energy well?” is subtle in wording but profound in practice.
In Part Four, I will turn specifically to leaders - and the challenge of regulating your own nervous system while supporting others through uncertainty.
References
Bosshard, M., & Gomez, P. (2024). Effectiveness of stress arousal reappraisal and stress-is-enhancing mindset interventions on task performance outcomes: A meta-analysis of randomized controlled trials. Scientific Reports, 14, 7923. https://doi.org/10.1038/s41598-024-58408-w
Brooks, A. W. (2014). Get excited: Reappraising pre-performance anxiety as excitement. Journal of Experimental Psychology: General, 143(3), 1144–1158. https://doi.org/10.1037/a0035325
Hase, A., Nietschke, M., Kloskowski, M., Szymanski, K., Moore, L., Jamieson, J. P., & Behnke, M. (2025). The effects of challenge and threat states on performance outcomes: An updated review and meta-analysis of recent findings. EXCLI Journal, 24, 151–176. https://doi.org/10.17179/excli2024-7995
Jin, H., & Peng, Y. (2024). The impact of team psychological safety on employee innovative performance: Communication behaviour as a mediator. PLOS ONE, 19(10), e0306629. https://doi.org/10.1371/journal.pone.0306629
Seery, M. D. (2011). Challenge or threat? Cardiovascular indexes of resilience and vulnerability to potential stress in humans. Neuroscience & Biobehavioral Reviews, 35(7), 1603–1610. https://doi.org/10.1016/j.neubiorev.2011.03.003
