Generative AI tools are making designers faster. They are also making design systems decay faster. And most teams have not yet noticed the second problem.
The output looks right. It just is not right. And the gap between "looks right" and "is right" is where design systems go to die.
The decay pattern
Design systems do not fail dramatically. They fail incrementally. A new designer joins the team and uses a generative tool to mock up a new feature. The tool produces components that look like the design system. They ship. Six months later, the product has a dozen variants of a button that were all generated, not governed.
The traditional fix is more documentation. More guidelines. More Figma page rules. More Confluence pages that nobody reads.
More documentation does not solve the problem. The problem is not that designers do not know the rules. The problem is that the rules are not present at the moment the decision is made.
What governance actually needs
Effective design governance needs to operate at decision time. Not before, when nobody knows what they will need. Not after, when the damage is already in the product. At the moment a designer is about to do something that conflicts with the system.
This is a different kind of tooling. It is not a documentation platform. It is an intelligence layer that understands the system, monitors for drift, and surfaces context when new decisions happen.
It also needs institutional memory. The reason the primary button is that shade of terracotta and not a brighter red is buried in a Figma comment from 2021 that nobody can find. If that reasoning is not captured and surfaced, new team members will override it without knowing they are overriding anything.
Why we are building Tenet
We built Tenet because we kept running into this problem with our clients. Design systems we had built carefully would drift within six months of handoff. Not because the teams were careless — because the systems had no way to defend themselves.
Tenet starts inside Figma, where the decisions are made. It scans for violations, surfaces the reasoning behind constraints, and captures new decisions as they happen. It is not documentation. It is institutional memory that knows when to speak up.
We are still in early development. If you are dealing with design system governance problems — especially in a team using generative tools — we would like to talk.