05/15/2025
AI and WordPress aren’t quite playing nicely yet—and it’s starting to show. We’re seeing AI reshape how people work across design, development, and content—but WordPress still feels awkward when it comes to real-world AI integration.
From a developer and project lead perspective, here’s the problem:
The current wave of “AI for WordPress” plugins mostly slap a ChatGPT textarea onto the editor and call it a feature. There’s very little real workflow support, no semantic awareness of your site structure, and definitely no native API for AI in core. It's all reactive, not foundational.
Some practical pain points:
- No structured AI assistance in Gutenberg (imagine AI generating custom blocks with real context?)
- No way to train models on your own post types, taxonomies, or content history
- Limited or nonexistent integration with ACF, CPT UI, or actual WP_Query logic
- Server performance becomes an issue when devs try to hack in AI features without offloading to real processing layers
What WordPress needs isn’t just more AI plugins—it needs a framework-level rethink.
Imagine:
- AI hooks and filters baked into core
- A standard schema to let AI understand custom content types
- Built-in AI queues or offloading pipelines
- Developer-first SDKs to let us control prompt flow, context windows, and post-processing
Right now, we’re just duct-taping OpenAI to WordPress—and calling it progress. But if WordPress wants to stay dominant in the CMS space, especially for content-heavy sites, we’ll need a real AI-native approach, not just more UI sugar.
Are you working on anything in this space? Anyone figured out real AI+WP architecture that isn't just a wrapper?
Let’s talk. Genuinely curious who’s agreeing, or who has found a solution for this?
Author: Hanna Partridge