top of page

What to expect, when everything Depends on Everything Else in cybersecurity

11 minutes ago
4 min read

19 November 2026 Joel Van Dyk | Joel Van Dyk Advisory, Ltd.


Figure 1: How Small Infrastructure Changes Create Cascading Effects
Figure 1: How Small Infrastructure Changes Create Cascading Effects

Modern digital infrastructure is built on interconnected systems, where even a routine software update, network change or certificate replacement can have consequences far beyond its original scope. For financial institutions, this is particularly important as they prepare to transition to post-quantum cryptography. Successful migration requires understanding these dependencies, testing changes in controlled environments and introducing new security measures without compromising operational resilience.

Source: Joel Van Dyk Advisory, Ltd. | November 2026

There are weeks when a story catches my attention because it is directly about quantum security. This wasn't one of them.


Over the past week, we've seen reports of disruptions affecting aviation, Microsoft services, Zoom, banking, national security, and other parts of the digital infrastructure we rely on every day. The individual circumstances differ. We should be careful not to assume that unrelated incidents share a common cause, or that every outage is a cyberattack, but taken together, they raise a familiar question.


Do we really understand how much of our infrastructure depends on systems outside our control?


The problem isn't always where the failure occurs

Most organizations have invested considerable time and money securing their own environments. They have redundancy, monitoring, incident response procedures and disaster recovery plans. What is less clear is how well those plans account for dependencies that sit beyond their immediate control. A service provider experiences an outage. An authentication platform becomes unavailable. A software update creates problems across thousands of systems.


Usually, we build in multiple redundancies into systems that touch critical processes such as banking. Many are the nights I remember as a trainee where we "canned" an upgrade/changeout that had been thoroughly tested, yet still didn't work out the way we had planned. Also frequent are the people we "beeped" (yes, this was the '90s) out of bed beause their cutover procedures were unclear or had failed or their system just went south. Changes were done and planned with multiple redundancies, multiple backups of people to escalate to, and done only at low usage times when markets were closed. For us in banking this was only Friday and Saturday nights.


Systems back then were much simpler and certainly not as interconnected as they are now. While that has changed, the approach to change control (emphasis on control) has not. Contain the change, be able to back away from the change if needed to get back to the last known good condition, because we had drummed into us that humans make mistakes. What is the fallback procedure and how fast can you recover? That is still the operational procedure these days. This also applies in spades to what happend in the last few days. I'm sure my collegues in the affected industries practice the above change controls daily, and so, why we still remain somewhat skeptical about explanations of maintenance errors. Also, while the original failure may be relatively contained, because of the interconnectivity, the consequences are not. Again, this is particularly relevant in financial services, where applications, payment systems, identity services, communications networks and third-party providers are heavily interconnected. A failure in one part of that infrastructure can have consequences well beyond the organization where it originated. This was drummed into us in the old days, and is even truer now.


Complexity has become a security problem

We've spent decades connecting systems to make them faster, more efficient, but also easier to manage. There are obvious benefits to that approach, but every connection also introduces a dependency, and every dependency needs to be understood. The challenge is no longer simply whether an individual system is secure. It's whether the wider environment can continue functioning when something it relies on becomes unavailable, that requires a different level of architectural understanding, it means knowing which services are critical, where the dependencies exist, who controls them and what happens when they fail. Part of this failure analysis is what happens if the security controls fail?


None of this is new, and something we all practice. It's the scale of the problem continues to grow.


And then there is quantum

Quantum computing introduces another dimension to this challenge. The transition to post-quantum cryptography is often discussed as a matter of replacing vulnerable algorithms with quantum-resistant alternatives.


Technically, that is part of the process.

Operationally, the challenge is considerably broader.


Cryptography is embedded throughout modern infrastructure in the same way a building is supported by its foundation. It supports authentication, communication, software distribution, financial transactions and the relationships between systems. Changing those mechanisms means understanding where they are used, how they interact and what depends on them. In complex financial environments, that is not a trivial exercise. Organizations will need to manage the transition while maintaining availability, confidentiality, integrity, and the security controls they already depend on.


The answer is to introduce new technology, but continue to analyze our understanding its operational consequences.


Resilience needs to be designed, not assumed

The lesson from recent disruptions is not that technology has suddenly become unreliable. It's that interconnected systems can produce consequences that are difficult to anticipate, even when the individual components appear well managed. Is supect this may be close to the explanation of this past weekend.


The same principle applies to quantum readiness. Before organizations start making significant changes to their cryptographic infrastructure, they need a clear understanding of what they have, how it is connected and what happens when something changes. That means treating cryptographic migration as an architectural, governance, and operational program, not simply a technology replacement project.


For financial institutions, the priority should continue to be to identify critical dependencies, understand the risks and establish a transition approach that can be tested without unnecessarily disrupting existing operations.


The bigger question

We tend to measure security by how effectively we prevent something from going wrong, but resilience is also about what happens when it inevitably does, and how fast we can recover. Can the organization identify the problem? Can it contain the impact? Can critical services continue operating? And does it understand the dependencies well enough to recover?


Those are the questions that matter during any change, including the past few days. As we move toward a post-quantum environment, they will become even more important, because protecting individual systems is only part of the job.


Lesson: we have to understand what happens when everything depends on everything else.


 
 
 

Comments


bottom of page