WordCamp US 2026 Has an AI Track. Here's What "Preparing Your Site for AI Agents" Actually Means

There is a line on the WordCamp US 2026 tickets page that most people will read straight past.

The AI track, they say, covers AI-assisted development, preparing your site for AI agents, and weighing the legal and ethical uncertainties.

That middle phrase is doing an enormous amount of work for four words. It is also the part of my job that most WordPress owners have not started thinking about yet. So let me unpack what it actually means, because if you are deciding whether that track is worth your time, you should know what is behind the label.

What is the AI track at WordCamp US 2026?

It is one of four tracks in this year's programme, and here is how WordCamp US describes it in their own words.

Alongside it sit a technical track on modern workflows and automation, a business track on pricing, maintenance and how the agency model is changing, and a newcomers track of approachable sessions.

Note the word they chose: practical. Not "the future of AI", not a keynote about disruption. Sessions about doing the thing.

That distinction matters, because most AI content aimed at WordPress owners right now is neither practical nor honest. It is a plugin advert wearing a thought-leadership hat.

WordCamp US 2026 runs 16 to 19 August at the Phoenix Convention Center. Sessions are on the Tuesday and Wednesday.

What does "preparing your site for AI agents" actually mean?

Start with what an AI agent is not. It is not a person. It does not scroll, it does not wait for your carousel, and it will not click your cookie banner.

An AI agent arrives at your site to extract an answer. Sometimes it is a crawler building an index. Sometimes it is a live fetch because someone asked ChatGPT a question this second and it is reading your page to reply. Increasingly it is something acting on a user's behalf, comparing options or filling something in.

Preparing for that comes down to three unglamorous questions.

Can it read you at all? If your content only appears after JavaScript runs, a good share of agents see an empty page. This is the single most common failure I find, and it is not an AI problem. It is an old technical SEO problem wearing new clothes.

Can it tell what you are? Entity clarity, clean structure, headings that match real questions. An agent quoting you needs to know what it is quoting and who said it.

Do you want it there? This is the one people skip, and it is a business decision, not a technical one.

That is the whole of it. No plugin installs this. It is the same technical SEO work you should already be doing, pointed at a reader who never became a visitor.

Should you block AI agents or let them in?

This is where the honest answer costs something.

If your traffic is your product, ad revenue, affiliate clicks, a content site that lives on sessions, then an AI agent reading your page and answering the question elsewhere is a straight loss. You did the work, someone else served the answer. Blocking AI crawlers is a legitimate response, and I have set it up for people who needed it.

But if being cited is the product, and for most service businesses it is, blocking removes you from the answers your buyers already trust. You cannot be named in a room you refused to enter.

The trap is treating it as one switch. It is not. Different agents do different jobs, and the crawler that trains a model is not the crawler that fetches your page to answer a live question. You can allow one and refuse the other.

Most sites I look at have never made this decision at all. They have simply inherited whatever their host or their robots.txt happened to say.

Where does the traffic go when nobody clicks?

Here is the part that unsettles people. A growing share of the value your site produces will never show up in your analytics.

Someone asks an AI a question, your page supplies the answer, they act on it, and you see nothing. No session, no referrer. I have written about this as AI dark traffic, because that is what it feels like when you first spot it: work leaving the building unmeasured.

This is why "did our traffic go up" is becoming the wrong question. The right one is whether you are being named when it counts, which needs different tools and a different definition of winning.

Bring that question to Phoenix and it stops being theoretical. There are people in that building who have measured it on sites much larger than yours.

Why the legal and ethical session is the one I would sit in

The third piece of the AI track is the one nobody markets, and it is the one I find most interesting.

Everything above is answerable. Can agents read your page, should you let them, how do you tell. Those have answers, and I can give you most of them in a paragraph.

Whether you should let a model consume work you paid a writer for, and what it means that your content trains the thing that may replace the search result you used to rank in, has no settled answer. Nobody has one. That is what "uncertainties" is admitting, and I respect that they used the word instead of pretending otherwise.

A room of people arguing about that in good faith is worth more than another blog post announcing that AI changes everything. You cannot get that from a forum thread, and you certainly cannot get it from an AI.

Is the AI track worth the ticket?

If you run WordPress sites and you have not yet decided what you are doing about agents, yes.

Not because the sessions will hand you an answer. Because the questions are the sort you only sharpen against other people. I have spent two years working on getting brands surfaced inside AI answers, and the things that moved my thinking came from arguments, not articles.

General admission is $100, or $80 with code AF26. Student tickets are $25. The whole thing runs 16 to 19 August at the Phoenix Convention Center.

The link applies the code for you, so the ticket reads $80 before you reach payment.

Final take

The AI track exists because WordPress is having the same argument the rest of the web is having, and Automattic has put it on a schedule instead of leaving it in the hallway.

My honest position has not changed in two years, and I will say it here as plainly as I say it to clients. GEO and AEO are not new disciplines. They are entity clarity, structured data and topical authority, the same fundamentals, pointed at surfaces that did not exist five years ago. The sites getting cited inside AI answers today are overwhelmingly the ones that got their basics right years ago.

That is either reassuring or annoying, depending on how much work you skipped.