Imagine a pack of monkeys breaking into your organization’s data center. They pull out wires, knock over servers, and create complete chaos. How would your systems fare — and how would you as an organization react?

This thought experiment is exactly what Netflix did when they introduced their Chaos Monkey test in 2011. They wanted to test how their own infrastructure, and organization, would function if a large number of the company's computers suddenly stopped working. The goal is to design systems that are prepared for attacks from chaotic monkeys, even though no one knows when they will arrive or what they will destroy. 

The result was a system that has become a benchmark for how we create real resilience, but Chaos Monkey is not really about monkeys or servers. It is about how people and organizations react when something unexpected happens. Because ultimately, we want to create anti-fragile systems. Anti-fragile systems are not just robust — they are improved by disruption. Each incident makes them stronger than before. Much like a shinbone gets stronger when it heals after a fracture. Then system architecture is not enough. It is also about how everyone in the organization behaves. So it is not just a technical issue, but also a social one. 

Let's return to the monkeys in the Netflix server room. If the organization's main focus becomes who who forgot to lock the door, we not only lose valuable time, we also risk thwarting learning. The main question – once we have secured operations – is of course what we can do to prevent it from happening again. Then we also need to have a culture where everyone dares to be open about their mistakes, and raises their hand to contribute. If I was the one who didn't double-check that the door was closed, I do the most good by telling you about it and perhaps pointing out that maybe the door closer isn't working properly. If I wrote the code that created a vulnerability, I contribute best by sharing how I thought and why it went wrong.  

This requires a safe culture that permeates the entire organization. To create and maintain one, individuals, teams, managers and employees in other parts of the company must take responsibility. Everyone needs to contribute to an organization where we not only do our best to maintain resilience and safety, but also where we all dare to talk about our failures. Otherwise, we cannot become stronger from incidents. Otherwise, we are not a learning organization. 

So my checklist after an incident is:

1. Create forums for learning — Document, discuss, and follow up on incidents systematically.
2. Ensure psychological safety — Ask yourself if people dare to talk about mistakes, hesitations, and observations.
3. Analyze behaviors and incentives — How do you react? How does the team react? How does the organization react? What drives these behaviors?

As I and my co-authors Daniel Deogun and Dan Sawano writes in the book Secure by Design Security is not just about code and architecture, it's just as much about creating a security-conscious culture. That's how we work at Omegapoint, and that's the gospel we spread to our customers, and to anyone else who will listen.