Failure Tests Technology.
Response Tests the Institution.

Last month I asked what happens next when trust breaks. This month I want to consider a harder case: what happens when the event was not the institution's fault at all?

Organizations invest enormous resources in preventing things from going wrong. They strengthen cybersecurity, establish controls, protect privacy, manage risk, and increasingly govern emerging technologies such as AI. These investments are essential.

But some events are partly—or substantially—outside an organization's control. A third-party provider may fail. External fraud may affect customers. A sophisticated attack may succeed despite strong defenses.

In those moments, customers may care about more than where responsibility ultimately lies. They face a more personal question:

The Cause May Not Be Yours. The Response Still Is.

When an event originates elsewhere, it can be tempting for an institution to view it primarily as someone else's problem: “The provider failed.” “The fraudster acted.” “The weather intervened.”

That may accurately describe the cause. But it may not fully describe the customer's experience.

An airline cannot prevent a storm. A bank cannot prevent every attempted fraud. A business cannot control every disruption involving one of its providers. Yet our view of these organizations can still be influenced by what they do when the normal experience breaks down.

Difficult Moments Reveal the Relationship

Every organization says it values trust. Difficult moments show what that commitment means when it is tested.

When everything works, trust is easy to claim. When something goes wrong—especially something the organization did not cause—customers notice how the institution behaves.

The stakes are also rarely equal. For the institution, an event may be one matter among many being handled by specialized teams. For the individual, the same event may affect something deeply personal or important in their life.

That difference in perspective matters.

Helping Should Not Mean Deciding for People

When something has gone wrong, there is a natural instinct to act quickly on the customer's behalf. Often that instinct is well intentioned.

But helping people should not mean unnecessarily taking decisions away from them.

An event that disrupts someone's life may already have reduced their sense of control. The response should not compound that experience.

People should be able to understand what is happening to them. They should have meaningful choices where choices are appropriate. Their privacy should be respected.

Trust Is Demonstrated Through Behavior

A thoughtful response cannot erase what happened. Nor should an organization assume that responding well will automatically restore confidence.

Customers make that judgment for themselves.

But what an institution does afterward gives them important information on which to make it.

As AI becomes more deeply embedded in consequential interactions, institutions will increasingly face difficult moments involving complex technologies and decisions. The standard for how they treat affected people should not be lower simply because technology was involved.

Organizations should continue investing aggressively in prevention, security, privacy, governance, and accountability.

But they should also recognize a simple reality:

Not every failure can be prevented. Not every event will be within an institution's control.

What happens afterward still matters.

Organizations cannot control every event.

But they can influence how customers experience what comes next.

And that experience may help determine whether trust ends with the event—or has an opportunity to continue beyond it.