There are very few management ideas that have genuinely changed how organisations approach culture. Psychological safety is one of them. When Harvard Business School professor Amy Edmondson defined it as ‘a shared belief that the environment is safe for interpersonal risk-taking’, she gave leaders and teams something that had been missing from the performance conversation for decades: a language for the relational conditions that make real work possible.
Google’s Project Aristotle – the company’s landmark research into what makes teams effective – reinforced the same principle; but this time at scale. Across hundreds of teams, psychological safety was the single most important factor determining team performance; not talent distribution, not process quality, not individual expertise, but the degree to which team members felt safe to take interpersonal risks.
This was transformative. It redirected a generation of leaders’ attention toward culture, toward cultivating openness, and toward the quality of the relational environment in which work gets done. Organisations invested in building psychological safety, leaders modelled vulnerability, and the conversation changed.
And Yet: Performance Still Stalls Under Pressure
And yet…
More than twenty years on, teams that have built genuine psychological safety are still stalling under pressure. They are still fracturing when conditions shift rapidly and still noticing the gap between what people feel in a well-facilitated workshop and what actually happens in a high-stakes Monday morning meeting or during a must-win close game.
The concept did not fail us. But we stopped too soon. And the way psychological safety has been framed and then applied in most organisations has created three specific misconceptions that are now limiting performance in precisely in those moments where it is most required.
Misconception #1: Treating Safety as Binary
The first misconception is treating psychological safety as binary. In most organisational conversations, the concept of safety is treated as though it is either present, or it is not – a threshold you cross, a state you achieve. But safety is not binary. It exists along a continuum, and within almost every team there are dynamics that cultivate greater safety alongside others that simultaneously sustain fear-based patterns. Treating it as binary creates a false sense of completeness and minimises the nuanced nature of how safety is felt, shared, and shifted.
Misconception #2: Assuming Safety Stays Built
The second misconception is assuming that safety, once built, stays built. In many organisations, a psychological safety initiative launches, metrics improve, and attention moves on. But safety is not a destination. It is a dynamic. New members join with different histories and different thresholds. Competitive pressure intensifies. A leadership change alters the relational landscape in ways that ripple through the entire team. Each of these events has the potential to erode safety rapidly – not through dramatic incidents but through the accumulation of small moments. An eye roll. A question that goes unanswered. A pattern of who gets interrupted and who does not. Safety fluctuates in real time, interaction by interaction. This is because what shifts psychological safety is primarily people’s behaviours, and then how other people interpret, make associations with, and respond to those behaviours. Treating it as stable is how organisations invest in culture and then lose it without knowing why.
Misconception #3: Confusing Safety with Comfort
The third misconception is to equate psychological safety with comfort. Psychological safety fosters openness and risk-taking in groups by increasing tolerance for discomfort. In contrast, comfort prioritises avoiding discomfort, thereby maintaining the status quo by inhibiting growth and innovation.
While comfort aims to foster tranquillity and consistency, psychological safety promotes growth through vulnerability and constructive feedback, thereby enriching collaboration and individual development.
I’ve seen teams where ‘safety’ becomes weaponised when it is treated as synonymous with comfort. In those environments, people can sidestep the rigour of growth by claiming they don’t feel safe – stalling progress and holding the rest of the team hostage, when what they’re actually experiencing is the natural discomfort that accompanies learning, change, and development. Leaders, therefore, need to be able to discern between the unavoidable discomfort that accompanies growth and the reactions that arise due to a genuine insufficiency of safety.
When safety is mistaken for comfort, teams can become polite but stagnant. People hold back honest feedback, avoid productive conflict, and settle for consensus over truth. In place of accountability and productive friction, complacency and placation progressively become the cultural norm. In this state, learning and innovation slow, and performance plateaus. Growth, by definition, is uncomfortable, and teams that understand this distinction are far more likely to optimise performance by cultivating the psychological safety necessary for them to become increasingly comfortable with the uncomfortable.
The Deeper Limitation: An Individual Framing
But perhaps the deepest limitation lies not in any of these three misconceptions. It lies in the fundamental framing of psychological safety as an individual experience. When safety is defined as whether “I feel safe to speak up”, responsibility for culture can inadvertently rest on the willingness of individuals to take interpersonal risks – which places the burden on the people who are most exposed, while leaving the system that shapes everyone’s behaviour largely unexamined. Organisations end up encouraging people to find their voice without addressing the environment that made speaking up feel dangerous in the first place.
A Necessary Reframing: A shift from psychological safety to Adaptive Safety
What is required is a reframing. Not a replacement – psychological safety remains a vital foundation – but an extension. A shift from asking ‘do people feel safe?’ to asking ‘how are we, together, creating or eroding safety right now, in this interaction, under these conditions?’
Adaptive Safety is the shared capability of a team to continuously create conditions of trust, connection, and inclusion – adjusting together as contexts and challenges evolve. It makes safety active rather than passive. Collective rather than individual. Dynamic rather than fixed. And it connects safety directly to the learning and adaptability that organisations in a fast-changing world cannot afford to live without.
Reframing Safety; Reframing Leadership
What can leaders do to shift how we perceive safety from an individual experience to collective responsibility?
First, notice patterns and dynamics that are consistent throughout the system. Separate these from individual psychology, capability, and work ethic, and rather, see them as windows into where the cultural infrastructure is resilient, and where it is under strain.
For example, if the same two voices dominate every meeting, that is not about confidence alone; it is a pattern of airtime distribution. If certain ideas consistently only surface after meetings, that is not about introversion; it is a signal about when and where it feels safe to contribute. Treat these patterns as data about the system, not deficiencies in individuals.
Second, make the interactional rules of the team explicit and shared. If safety is collective, then the behaviours that sustain it cannot remain implicit or assumed. Leaders need to define, with the team, what productive challenge, dissent, and contribution actually look like under pressure.
For example, a team might agree that interruption is acceptable only to build, not to dismiss. Or that the role of the most senior person in the room is to speak last, not first. Or that disagreement must be made visible in the room, not deferred to side conversations. These are not soft agreements; they are operating standards that shape how risk is distributed across the group.
Third, intervene in real time when the system drifts. Culture is not set in workshops; it is maintained or eroded in moments. Leaders must shift from being observers of behaviour to active stewards of the environment as it unfolds.
For example, when a contribution is overlooked, bring it back into the room and give it weight. When tension rises and the group defaults to politeness, name what is not being said and re-open the conversation. When someone takes a risk and is subtly shut down, pause and recalibrate the interaction before moving on. These micro-interventions signal that safety is not the responsibility of the individual who spoke up, but of the system that receives them.
When leaders make this shift, psychological safety evolves from something people are asked to feel, into something teams are required to create. And it is in that shift – from individual courage to collective responsibility – that safety becomes not just protective, but truly performance-enabling under pressure.
Going Further
Psychological safety told us that people perform better when they feel safe to take risks. That insight is as true now as it was twenty years ago. Adaptive Safety tells us how to sustain that safety in a world that keeps changing – and what it actually takes to move from a team that feels psychologically safe to a team that performs exceptionally under genuine pressure.
We did not stop too soon because the idea was wrong. We stopped too soon because the problem is harder than we thought. It is time to go further.
Next week, we turn to what I believe is the most expensive performance problem in most organisations; one that never appears on any dashboard. It lives in the idea that went unspoken in yesterday’s meeting, and the one that won’t be raised tomorrow either. Because what looks like underperformance very often isn’t. It’s fear. And until leaders can tell the difference, they will keep solving for the wrong thing.
From Fear to Flow: The Adaptive Safety Revolution by Errol Amerasekera is available now wherever books are sold. Each blog in this series draws on research and frameworks from the book – which goes substantially deeper on every concept introduced here.