The Future of WordPress: How AI is Becoming Part of the Platform

WordPress currently powers around 43% of all websites on the internet, according to W3Techs. Among sites that use a known content management system, that figure climbs to over 60%. Whether you look at the data through BuiltWith’s traffic analysis or W3Techs’ historical tracking, the conclusion is the same: no other platform comes close.

With that kind of scale, it was never a question of whether AI would become part of WordPress. It was always about how the project would approach it. And if you have been paying attention over the last year, the answer is becoming clear: carefully, transparently, and with community input baked into every step.

At the same time, there is a real conversation happening in the WordPress community about what AI means for the people who actually build and maintain sites. Developers are not being replaced – they are adapting, and the ones who engage with AI tooling early are finding it genuinely useful for the repetitive parts of their work.

In this post, we will look at what WordPress is actually doing with AI, unpack what the official AI Experiments plugin does today, and explore what this means for the workflows of developers, agencies, and content professionals over the next few years. We will also touch on how this shift connects to tools and themes that are already adapting to an AI-influenced workflow, including options like the Porto WordPress Theme that many developers already use for production sites.

How WordPress is Approaching AI Integration

First, it helps to separate two things that often get mixed together in conversation. On one side, you have the hundreds of third-party AI plugins already available in the WordPress directory – chatbots, AI content writers, image generators, and so on. On the other side, you have the official AI work happening inside the WordPress project itself. They are related, but they are not the same thing.

In May 2025, WordPress formally announced the creation of a dedicated AI team. Their stated goal was to explore and coordinate artificial intelligence projects across the WordPress ecosystem. This is a genuine effort to build a shared, open-source foundation for AI that respects how WordPress has always worked: through community consensus, open contribution, and gradual adoption.

Instead of jumping straight to flashy AI interfaces in core, the project is taking a foundational route. The WordPress AI team identified three building blocks that everything else will sit on top of:

1. PHP AI Client SDK:Provides a unified way to interact with large language models from different providers, including OpenAI, Google, and Anthropic.
2. Abilities API:Acts as a registry that describes what a given WordPress site can do (its “abilities”) in a format AI agents and automation tools can understand and call safely.
3. Model Context Protocol (MCP) Adapter:Translates those abilities into a standard protocol so external AI tools can discover and interact with WordPress sites in a structured way.

Together, these three components form the technical infrastructure for how AI will work inside WordPress going forward. They are not finished products. They are open building blocks, hosted publicly on GitHub, designed to be extended, tested, and refined by the community over time.

What the AI Experiments Plugin Actually Does

In late November 2025, the WordPress AI team released version 0.1.0 of the AI Experiments plugin. It is a canonical plugin, meaning it is developed and maintained by the WordPress project itself, not a third-party author. You can install it directly from the WordPress plugin directory.

The plugin is exactly what its name suggests: a safe, opt-in space for experimenting with AI-powered features inside the WordPress admin and editor. Nothing runs automatically, and nothing changes your content without your approval.

What Works Right Now

Right now, AI Experiments includes several opt-in features: title suggestions, excerpt generation, content summarisation, alt text suggestions, image generation inside the block editor, and an Abilities Explorer that lets you inspect and test registered abilities.

All of these run through the AI provider you configure (OpenAI, Google, Anthropic, etc.), using your own API key.

What is on the Roadmap

On the roadmap, the AI team has flagged experiments like contextual tagging, AI-assisted comment moderation, type-ahead suggestions while you write, logging and observability for AI requests, and a dedicated AI Playground for trying different models and prompts.

Some of these may shift or evolve as community testing continues, but the direction is clear: WordPress wants to make AI a native, structured part of content and admin workflows, not a bolt-on extra.

That honesty about what exists today versus what is planned is refreshing, and it sets a realistic tone for how AI will evolve inside the platform.

Why This Matters for WordPress Developers and Agencies

So what does all of this actually mean for the people who build, manage, and publish on WordPress every day?

Rather than treating each audience separately, it is worth looking at the four areas where this shift has the most practical impact – because they are all connected.

Opt-In Control and a Framework Built for Extension

The experiment framework is designed around choice. You enable only the AI features you want, pick your own AI provider, and decide whether to act on the suggestions or ignore them entirely. For agencies working across multiple client sites, this flexibility is essential. A small business in Western Sydney might need simple title suggestions to speed up their blog workflow, while a larger Australian enterprise client might want to build custom experiments tailored to their industry.

What makes this particularly useful for developers is that the plugin is not a black box. It is a reference implementation you can study and extend. There is an Abstract_Experiment base class for building your own AI features, hooks and filters for overriding prompts and responses, and the ability to pre-configure providers so clients do not need to manage their own API keys. This is the kind of custom website development thinking that makes WordPress so adaptable, giving you scaffolding to build exactly what each project needs, rather than a rigid, one-size-fits-all AI experience.

A Broader Ecosystem That is Already Adapting

While the official plugin takes a measured approach, the wider WordPress ecosystem has been moving fast.

The Porto WordPress Theme, for example, already supports advanced customisation through its builder tools, and themes built with this level of flexibility are naturally better positioned to integrate with AI-driven workflows as they mature.

As AI capability in core becomes more structured through the Abilities API, themes like Porto can start plugging into those patterns rather than building isolated features. You can explore Porto’s full range of templates at their official landing page. Porto has also been writing about the impact of AI on WordPress theme customisation, which is worth a read if you are thinking about how your theme choices today will hold up in an AI-integrated future.

Real Impact on Content and SEO Workflows

For content and SEO professionals, the impact of WordPress AI is very down-to-earth. It does not replace strategy or judgement, but it does take some of the grind out of everyday publishing.

• Faster headlines and summaries: Title suggestions, excerpt generation, and content summarisation give you solid starting points instead of blank fields, which adds up across a busy content calendar.
• Better alt text, more often: AI-suggested alt text at upload helps fix a common gap by making images more accessible and easier for search engines to understand.
• Cleaner taxonomy with less effort: As contextual tagging matures, AI-assisted suggestions for tags and categories can reduce manual taxonomy work and support a clearer site structure.
• Consistent quality across large sites: On bigger installs, these small assists help keep metadata, summaries, and image descriptions more consistent from one author or team to the next.

The Open-Source Advantage

There is a bigger picture here that is easy to overlook. WordPress is not a proprietary platform controlled by a single company, but an open-source project with a global community of contributors. When WordPress builds AI tooling, it does so in the open – public GitHub repositories, community Slack channels, and a transparent development process anyone can follow and contribute to.

This means the AI features that eventually make it into WordPress core will have been tested, debated, and refined by real users and developers before they ship. You are not at the mercy of a product team’s internal roadmap. You can file issues, submit pull requests, and help shape how AI works inside the platform you depend on.

For agencies and developers who build businesses on WordPress, that is not just reassuring – it is a genuine competitive advantage. Digital Sydney based agency, as well as other online marketing companies that stay close to platform-level developments are better equipped to advise clients on what is coming and how to prepare, which makes a real difference when you are helping businesses plan their digital strategy over the next two to three years.

Practical Next Steps for Technical Site Owners

If you are reading this as a WordPress developer, agency owner, or technical marketer, you do not need to overhaul everything overnight. But there are a few practical steps worth taking now:

Test AI Experiments in a Staging Environment

Spin up a staging site running WordPress 6.9 or higher, install the AI Experiments plugin, and connect it to a provider you are comfortable with.

Start with the low-risk stuff: use AI title suggestions on internal posts, see how the generated headlines compare to what you would write, and get a feel for how the tool behaves with your content and tone. If you are a developer, explore the plugin’s codebase on GitHub – it is designed to be studied and extended.

Map Your Workflows to Potential AI Abilities

Look at your current processes and ask: which steps could be expressed as clear abilities? “Create a draft product page from a structured brief.” “Review existing posts for outdated pricing.” “Tag new articles based on topic clusters.” If you can describe a task in those terms, it is a good candidate for future AI-assisted work via the Abilities API or custom experiments. Even if you do not build those integrations immediately, you are preparing your architecture and mental model for them.

Decide What You Actually Want AI to Do

It is easy to get carried away with possibilities. The more sustainable path is to decide what you actually want AI to handle: reduce repetitive manual work, improve accessibility and metadata quality, support editors who are not native English speakers, or speed up testing of different headline and description variations.

Once you know your priorities, you can choose which experiments to enable and where to invest development time, instead of turning everything on and hoping for the best.

Bring Development and Marketing Together Early

Many of the long-term benefits of AI in WordPress sit at the intersection of development and content strategy: content modelling, SEO, data, user journeys. That is why agencies that already combine these skills under one roof are in the best position to take advantage of what is coming.

If your team is lean or smaller, it may simply mean putting your developers, SEO specialists, and content people in the same planning conversations when you discuss AI features rather than treating AI as a pure developer toy or a pure marketing shortcut.

Where WordPress AI is Headed from Here

If you are wondering what the AI landscape inside WordPress will look like by late 2026 or into 2027, the honest answer is that nobody knows for certain. But we can read the signals clearly enough.

The AI Experiments plugin is a foundation, not a finished product. Its architecture with the Abilities API, the WP AI Client SDK, and the MCP Adapter is designed to support far more than title suggestions and excerpt generation. The WordPress AI team has talked about smarter authoring and editing experiences, richer media tools, and more adaptive admin interfaces. These are not vague promises; they are grounded in technical infrastructure that already exists and is publicly available.

What is likely to happen is a gradual process where successful experiments migrate from the plugin into WordPress core, following the same pattern that Gutenberg used. Features will be tested, iterated on, and eventually stabilised. The pace will depend on community feedback and real-world usage data.

For now, the tools are being built, the foundation is in place, and the community is open to input. That combination has always been WordPress’s greatest strength. And it is exactly what responsible AI integration needs to succeed.