When WordPress announced native AI connectors for providers like Anthropic, OpenAI and Google, I immediately wanted to test it.
Because honestly: this is a pretty significant moment for WordPress.
For years, AI workflows around content creation have mostly lived outside the CMS itself. You would do your research in one tool, prompting in another, optimization somewhere else, and then eventually move everything manually into WordPress.
Now WordPress is clearly trying to change that.
So I went through the new AI functionality in WordPress 7.0, connected different providers, tested the available features, and tried to answer one simple question:
Is this already genuinely useful, or are we mostly looking at an early foundation for something much bigger?
After testing it hands-on, I think the answer is both. Let’s have a look:
You are currently viewing a placeholder content from YouTube. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationWhat WordPress actually introduced
The biggest change in WordPress 7.0 is the introduction of the Connectors API.
In simple terms, WordPress now has a centralized system for connecting external AI services directly into the CMS. Right now, the supported providers are:
- Claude via Anthropic
- ChatGPT via OpenAI
- Gemini via Google
The entire system lives inside a new Settings → Connectors area in WordPress, where users can configure API keys and manage providers directly.
Technically, I actually think this is a very important step.
Because this is bigger than just “AI features inside WordPress.”
This is WordPress officially introducing AI infrastructure into its core ecosystem.
That matters.
What you can do with it today
Once connected, WordPress gives you several AI-powered editing capabilities directly inside the editor.
That includes things like:
- Title generation
- Excerpt generation
- Meta description generation
- Content summarization
- Editorial notes
- Content classification
- Text shortening
- Text expansion
- Rephrasing
- Basic image generation
At first glance, this feels exciting. And to be fair, for many casual WordPress users this probably is useful already.
But after testing it properly, I quickly ran into the biggest issue.
The problem: the AI feels too rigid
The current implementation gives you AI functionality, but not much control.
For example, when regenerating titles, WordPress changed capitalization structures automatically without understanding the editorial style I originally used.
And the same thing happens with the shorten, expand, and rephrase features.
They technically work.
But they operate behind locked prompts with very limited contextual control.
As somebody working deeply in SEO and AI-assisted publishing workflows every day, this becomes noticeable immediately.
Because good AI output is rarely about pressing a button.
It is about context.
You need to guide tone of voice, brand positioning, audience expectations, SEO intent, topical depth, internal linking logic, formatting preferences, and content strategy itself.
And right now, WordPress still feels very far away from that level of intelligence.
AI features are not the same as an AI workflow
This is the key difference I think many people will miss.
Generating a title is not a content workflow.
Rephrasing a paragraph is not a publishing system.
Real content operations require an entire pipeline:
- Research
- Search intent analysis
- Topic validation
- Outline generation
- Content structuring
- Metadata optimization
- Image creation
- Internal linking
- Performance measurement
- Continuous optimization
That full cycle still does not exist natively inside WordPress.
And honestly, this is why many AI-powered SEO platforms already feel significantly more mature than what WordPress currently offers out of the box.
Because the real challenge was never generating text.
The real challenge is orchestrating the entire workflow around it.
Why I still think this update matters a lot
Despite the current limitations, I still believe this is one of the most important WordPress updates in years.
Not because of what it does today.
But because of what it enables tomorrow.
WordPress is no longer treating AI as a third-party experiment. It is slowly becoming part of the CMS architecture itself.
That changes the direction of the ecosystem.
Once these connectors become more deeply integrated, the possibilities become extremely interesting.
Imagine a future where:
- Claude researches and structures topics directly inside WordPress
- ChatGPT generates complete publishing drafts with contextual SEO logic
- Gemini creates supporting media assets automatically
- AI agents optimize internal links dynamically
- Metadata, schema, excerpts, alt text, and slugs are generated contextually
- Performance data continuously feeds optimization recommendations back into the system
That is where this becomes truly powerful.
And honestly, I do not think we are that far away anymore.
The real missing piece: connected AI systems
Right now, most advanced AI publishing workflows still happen outside WordPress.
People use ChatGPT, Claude, Gemini, APIs, automation tools, MCP connectors, or custom systems to create end-to-end content pipelines.
The average WordPress user cannot realistically build these systems themselves.
That is exactly why this update is so important.
Because WordPress is laying the groundwork for AI-native publishing experiences that normal users can eventually access without technical complexity.
And when that moment comes, content operations inside CMS platforms will fundamentally change.
My final verdict
After testing the WordPress 7.0 AI Connectors extensively, my conclusion is relatively straightforward:
The infrastructure is exciting.
The current execution still feels early.
If you already work with advanced AI SEO or AI-assisted publishing workflows, most of the current functionality will not feel revolutionary yet. Similar features have existed for years inside specialized SEO tools and AI publishing systems.
But WordPress moving AI directly into its core architecture is a major signal for where the web is heading.
This is probably not the final form of AI inside WordPress.
It is the foundation layer.
And over the next years, I genuinely think this could evolve into one of the biggest shifts in how websites are built, managed, optimized, and scaled online.