Leaders need to understand how AI is interacting with psychological safety in their organizations. When someone uses AI to surface a concern without owning it, this is called “safety by proxy.” This phenomenon occurs when people use AI’s perceived neutrality to raise concerns they don’t feel safe enough to raise directly.
This can be a signal that something important is going unsaid, and leaders should pay attention to it. AI’s perceived neutrality can also work as an equalizer, allowing underrepresented team members to route concerns through AI and have them arrive without a name attached.
Understanding the Patterns
There are four distinct ways AI’s perceived neutrality is interacting with psychological safety. If safety is absent, AI can serve as a workaround. If safety is unequal, AI can work as an equalizer. If safety is developing, AI can serve as a testing space. And AI can be used as a mirror, allowing leaders to gain access to input that their authority inhibits.
At 3M, the New Product Introduction process follows a rigorous stage-gate methodology, and the company’s culture has long held the belief that the best ideas can come from anywhere. The equalizer pattern is particularly visible in new product commercialization work, where established scientists use AI-assisted analysis to surface market perspectives.
Leaders’ Roles
Leaders should interpret AI-surfaced concerns as a signal, not a threat. They should evaluate the content on its merits, ask who benefits from the conclusion, and treat the channel as a reason for curiosity rather than as a verdict on the content’s validity. Leaders should also run their own stress-tests with AI, engaging with concerns before announcing a decision.
They should listen to people before they decide, and respond to acts of voice. When concerns surface, leaders should engage genuinely, acknowledge them, explain their reasoning, or adjust their position. This demonstrates that speaking up actually changes something. The goal is consistent, respectful engagement, including engagement that ends in a respectful no.
The window created by AI’s perceived neutrality will not stay open indefinitely. As AI becomes more organizationally embedded, its perceived neutrality will erode. Leaders who build genuine psychological safety now, while AI is still perceived as neutral, can create something durable. They will be able to establish a culture where team members feel safe to raise concerns directly, without relying on AI as a proxy.
