AI Governance: Without communication it is like politics without voters, beautiful in theory, but irrelevant in practice

HI recently read an AI governance framework that was impeccable: clear principles, defined responsibilities, well-structured processes. The problem is that it seemed designed to be read only by lawyers, technologists and regulators. No one else.

And that's a much more serious problem than it seems.

When design does not include communication, governance becomes invisible

I have been translating complex information for various audiences for 20 years. I have worked in public health, food security, environmental research, international development. And I learned something that governance frameworks often forget: people don't reject what they understand. He rejects what is not clear to him.

When I worked on behavior and social change, we discovered something forceful: health programs that failed did not do so because people were ignorant. They failed because no one had invested real time in explaining why something mattered, what would happen if nothing was done, and how the decision of each person affected the group.

Exactly the opposite of what we are doing with AI governance now.

AI regulation frameworks are being designed, many times, in technical and political spaces. They are well thought out. But when they go down to the field, to companies, to citizens, to local governments, they generate distrust. And that's not in bad faith. It's because of a lack of clarity.

A regulator can argue that a restriction on an AI model is "necessary to mitigate bias." But if that doesn't translate to "this means your company will be able to use AI without discriminating against your customers, and here's how," the message doesn't come in. It stays on the surface. And when it doesn't enter, it doesn't generate commitment. Generates obstruction.

Governance without social legitimacy is just a beautiful document

Here comes the uncomfortable part: the legitimacy of a governance framework does not come from how technically correct it is. It comes from enough actors understanding it, seeing the value, and being willing to implement it.

That requires communication. No translation of what already exists. Strategic communication from the design.

In international politics, we know that a policy is effective when:

People understand why it exists

People see that it affects their reality (positively or negatively)

People believe it was designed considering their interests

If one of those three is missing, politics may be perfect on paper, but the actors will find ways to get around it, criticize it, or directly ignore it.

Something similar is happening with AI governance. We have experts writing very sophisticated frames. But no one is asking: Does a small company understand what this means to it? Does a local government understand how this impacts its capacity for innovation? Do the citizens understand why this protects them?

If there is no clear answer to those questions, governance will face resistance. Not because it's bad. Because it will be perceived as opaque.

What is missing: strategic communication before, not after

Many governance initiatives expect to communicate after the framework is ready. "First we design, then we explain." It's like writing a novel without thinking about who is going to read it.

What works is different: thinking about communication from the beginning.

What does that mean in practice?

It means that when an AI governance framework is designed, the question is not only "is it technically correct?" But "how do we explain this to actors with different interests and understandings? What does a regulator need to understand, what does a technologist, what does an end user, what does an entrepreneur? What are the points of friction where clarity can change adoption?"

It means using narratives. Use examples. Use stories of risk and opportunity that people understand in their context.

It means recognizing that "AI governance" is not a technical concept. It's a social concept. And everything that is social requires clear communication.

And what would change if we started thinking like that?

Probably more robust frames, because they would face less resistance. Faster adoption, because people would understand the "why" before facing the "how". Less litigation, because there would be less room for conflicting interpretations. More legitimacy, because the actors would feel that their concerns were considered from the beginning.

But for that we need strategic communicators to be at the table when designing governance. Not later. During.

And that requires those who design governance to recognize something that remains controversial in technical spaces: communication is not a product. It's a process. It's thought. It's design.

Without that, you can have the best AI governance framework in the world. But it's going to be a document that people read, don't really understand, and then ignore.

Like a law that no one knows. As a policy that works perfectly on paper but that no one implements in reality.