When AI hit, part of me was devastated

I'll be honest: when AI hit, part of me was devastated.

Twelve years learning to build content systems that don't break. Permissions that mean something. Workflows where the wrong sentence doesn't just go live. And suddenly everyone's cheering about generating a thousand pages before lunch.

I thought: that was never the hard part. The hard part is content you can trust. Changes that go through review, not around it. In my world, a published mistake isn't embarrassing — it has consequences.

Then I watched what Drupal did with AI in 2026.

No chatbot bolted on the side. They split the work into two honest halves. Inside AI: the assistants and page-building that help people working within Drupal. Outside AI: making Drupal something external agents can safely connect to, inspect and change. Both built on the things Drupal was already good at — structured content, real permissions, approval workflows, versioning, so AI changes get the same scrutiny as human ones.

Nothing goes live unchecked.

The pitch isn't “make more”. It's “raise the bar, with governance underneath”.

After years of governance being the thankless thing nobody notices, it turns out that's the exact thing that makes AI safe to use.

I'm not devastated anymore. I'm a little proud. The future of AI in Drupal isn't a machine replacing the careful work. It's a machine that finally has to respect it.