Conway’s Law: What Your Culture Has to Do with Your AI

Have you heard of Conway’s Law? It is sort of like Murphy’s Law, except instead of everything that can go wrong going wrong, everything that is already wrong gets baked into the next system you build.
More precisely, Conway’s Law states that systems reflect the organization that builds them. In the beginning that is usually true. Teams design systems based on how modules interoperate, who can access which data, and where decisions and visibility lie. Over time the relationship begins to reverse. Once a system becomes central to daily operations, the organization gets accustomed to the flow of the system itself. Communication patterns form around it. Future systems and transformation efforts build on top of that structure and reinforce it again and again. What began as a system reflecting the organization gradually becomes the organization reflecting the system. This is about to happen with AI, and most organizations won’t see it coming.
That structure gets so locked in it persists long beyond every upgrade that follows.

Now you know why some approvals still route through a mailbox that feels like a historical artifact.

Nothing illustrates this better than the payroll system that once shaped how your organization communicated. The people who built it have retired. The vendor may no longer exist. The system itself may have been replaced more than once. And yet the structure it created often remains.
That payroll system did more than process paychecks. It became the operating rhythm of the organization. Decisions routed around it… teams organized themselves in relation to it. New systems were built to accommodate it, yet nobody chose this. It just stuck. Without deliberate intervention, your AI systems will do exactly the same thing.

Why It Matters Now

AI is revealing communication patterns and decision pathways that have been invisible for decades. It is also creating new ones that did not exist before. That combination is rare. Organizations that recognize this as a workforce design decision, not just a technology decision, can use this moment to build collaboration patterns and human AI teaming that reflect how they want to operate. Organizations that recognize this as a workforce design decision, not just a technology decision, can use this moment to build collaboration patterns and human AI teaming that reflect how they want to operate.

Getting your culture and human AI teaming right before the systems lock it in for you is what outsizes an AI investment. Conway saw this in 1967. The organizations that see it now have an extraordinary opportunity ahead of them.

Picture of Denise D'Angelo

Denise D'Angelo

Denise is a seasoned technology leader and mom of three who helps women lead change with clarity and confidence. These posts share her real-time thoughts on AI, transformation, and building sustainable success.

Share this Post:

Leave a Reply

Related Posts

Operating Models

Governance Isn’t a Gate. It’s a Game Plan.

Governance is often mistaken for a checklist that sits outside the work. In reality it functions more like Agile or Total Quality Management. Progress depends on people across the organization participating in how decisions are made, how risk is interpreted and how judgment is exercised while work moves forward. When governance becomes part of the rhythm of the team, coordination replaces compliance.

Read More

Discover more from dynamicD

Subscribe now to keep reading and get access to the full archive.

Continue reading