Personal AI agents akin Meta’s Muse and OpenAI’s Dots are splitting the web into two kinds of websites: the ones the delegate has a agreement with, and the ones it walks into pretending to be your customer.
Muse and Dots the two obtain a petition from a individual and transport it out on the web. They investigation a product, inhabit in a form, and can equal add the merchandise to cart and inspect out. Meta has documented how Muse does this, in detail. OpenAI has published much small concerning Dots, so most of this part is concerning Muse, alongside Dots anywhere OpenAI has stated something.
I think agents akin these volition increasingly be how group use the internet. Before the iPhone made mobile browsing a genuine thing, computers were the lone way most group always interacted alongside websites. Computers’ portion of traffic cut from 98% of worldwide web traffic in January 2010 to under fractional in 2026, by StatCounter’s count. We are concerning to see that identical item happen to browsers used by humans, as agent-driven browsers obtain a share.
The comparing, booking, and reordering a individual used to do throughout 10 tabs can now rotate into one request. Muse is already accessible to anyone 18 or complete in the U.S. and Canada, liberated for most uses. That is why this matters now, equal before anyone knows how many group shop this way.
For SEO teams, the part anywhere the AI delegate searches is acquainted ground, and so is most of what it says formerly it arrives: headings, nexus text, alt text, and the facts in apparent text. The part anywhere it completes the purchase, the named buttons, tagged fields, and checkout, has continually belonged to front-end teams.
An delegate in a browser finds websites the way a individual does, by name or by searching, so the ranking activity motionless applies, equal if neither Meta nor OpenAI says which hunt motor its delegate uses. After the delegate arrives, page structure decides whether it can peruse the page, engage alongside it, and complete the task, whether that is a booking, a form, or a purchase.
For marketers, attribution and spectators data assume the visitant is the person. When an delegate browses, your analytics volition most apt figure it as your customer. When it buys through a connector, your website analytics sees nothing, and the command appears lone in your platform’s command records.
This part covers which SEO controls motionless use to these agents, which ones no longer work, anywhere delegate action shows up in reporting, and what to inspect on your or your clients’ pages.
Muse And Dots Reach A Business Through A Connector Or A Cloud Browser
Muse’s archetypal tier is the connector. Meta says Muse “launched alongside dozens of partners alongside alongside admission to the complete Shopify catalogue,” and is adding Walmart, Best Buy, Sephora, Wayfair, and others. For the connectors it launched with, Meta says it “worked closely alongside the assistance provider to merge their API.” Dots have their own version. OpenAI says: “Through our ecosystem of plugins, they can readily nexus to complete 4,000 apps.” Any endeavor can build one and present it for OpenAI’s review.
The browser tier is everyone else. Muse runs, in Meta’s words, “a genuine up-to-date Chromium-based browser” on “a dedicated safe computer.” OpenAI says dots “can do nearly item using their own haze computer, their own browser, and the apps you’ve connected.” Meta’s documentation for businesses construction connectors says: “Muse can already browse websites. Your connector should enhance on what’s imaginable in the browser cognition through improved information, document access, additional dependable actions, or small steps.” Without a connector, the delegate comes to your website in its browser since a individual gave it a task, specified as purchasing a garment from a named designer, and it plant alongside any the leaf shows it.
If you’re on Shopify, the phase can choose your tier for you. Its default setting, “Allow Shopify to oversee for me,” is established under Sales channels, afterward Agentic, enrolls stores in its AI buying channels. Shopify says products are “shared alongside Meta by default through Shopify Catalog.” A shop on a tradition phase is in the browser tier unless it builds a connector.
I archetypal described these two tiers in Meta’s Muse documents in September.
See also: Getting Your Product Into ChatGPT Isn’t The Hard Part, Getting It Through Checkout Is
Muse And Dots Give Websites No Name To Put In Robots.txt
Most of the controls SEO teams use, from robots.txt rules to crawler allowlists, were built for crawlers that province who they are. Neither Meta nor OpenAI has published a person agent, petition header or signature for the Muse or Dots’ browsers, so there’s no name to put in robots.txt. Meta’s page for website owners lists five crawlers, and Muse is not one of them. OpenAI’s page lists four, and dots are not on that catalog either.
Meta describes what a website sees whenever Muse visits it: “When Muse browses the internet, it volition appear as your activity, so if you ask Muse to buy a garment from a garments designer’s website, that decorator power use your sojourn to display you an ad on Instagram.”
Think of Frank Abagnale in Catch Me If You Can. He walked through airports in a pilot’s uniform, and nobody checked whether he could fly, since the uniform was right. Muse walks into your website wearing your customer’s uniform and, in Meta’s words, “your activity.” Robots.txt lone always worked alongside names, and this visitant doesn’t provision one.
Robots.txt has continually been an regard system. The standard says its rules “are not a form of admission authorization,” and a crawler follows them lone if its controller chooses to.
A name lets a website acknowledge the visitant and decide what to do alongside it. For example, Amazon’s robots.txt names 99 agents, and Muse and Dots are absent since there is no name to write. When Amazon blocked Muse in September, it did it through its Conditions of Use, the conditions its customers concur to, as I wrote in Amazon’s block of Muse. GeekWire reported Amazon’s three reasons: Meta didn’t inform Amazon that Muse would admission its store, the delegate doesn’t acknowledge itself whenever it browses, and it appears to grasp and shop client credentials.
Dots have a second route. OpenAI’s documentation tells users: “If a website blocks the haze browser, ask your dot to try your connected computer.” For ChatGPT’s device use additional broadly, OpenAI says websites “may treat approved clicks, form submissions, and signed-in actions as coming from your account.”
For SEO teams, the decision moves from robots.txt to bot protection, and that is normally person else’s system. A firewall or your platform’s bot display sees a Chromium browser on a haze machine, and it either blocks it or lets it through as a person.
HUMAN Security, as reported by PPC Land, says its bot-detection merchandise can acknowledge Muse, which is a logic to ask your bot vendor whether it can inform if a Chromium browser is operated by a individual or an AI agent.
See also: Everyone Is Negotiating With Google While Meta Reads The Web For Free
On-Page SEO Covers Most Of What Personal AI Agents Read
Meta says Muse’s browsing sub-agent says a leaf through the accessibility tree: “The sub-agent, which drives the Browser, sees an accessibility tree snapshot of the page, not the raw DOM.” The sub-agent additionally “has no capability to run JavaScript in the leaf context.” Your page’s own scripts motionless run, since it is a genuine browser. Muse can’t add any of its own. OpenAI has not published how the dots browser says a page. Google’s guidance for website owners says browser agents may collect what they need by “analyzing ocular renderings (like screenshots), inspecting the DOM structure, and interpreting the accessibility tree.”
The accessibility tree is what display readers use. The browser builds it from the DOM and keeps, in Chrome’s description, “only elements from the DOM tree that are applicable and helpful for displaying the page’s contents in a display reader.” Each item has a category (button, link, heading), a name, and a state. Design does not appear in it.
Most of what the delegate says is on-page SEO you already do, but the part that completes the transaction is front-end work.
| Page element | In Chrome’s accessibility tree? | What it method for the agent | The check, and who normally owns it |
|---|---|---|---|
| Headings | Yes, alongside their level | The heading outline is the leaf outline | One H1, genuine heading tags in order, no styled divs position in for headings (SEO) |
| Link text | Yes, as the link’s name | “Read more” and “Click here” name nothing | Anchor content that says anywhere the nexus goes (SEO) |
| Image alt text | Yes, as the image’s name | A merchandise depiction alongside bare alt is an unnamed object | Alt content on all merchandise depiction (SEO) |
| Prices, sizes, and specs | Only as text | A cost in an depiction or a canvas has no content to read | Prices and options in HTML content (SEO, alongside front-end) |
| Buttons and icon controls | Yes, if they have an accessible name | An icon-only add-to-cart alongside no tag has no name | A apparent tag or aria-label on all authority (front-end) |
| Form fields | Yes, alongside their labels | Unlabeled size and color pickers are guesswork | A tag on all field, including tradition dropdowns (front-end) |
| Content in crumbled tabs and accordions | Depends on how it is hidden | A division crumbled by height stays readable. One hidden alongside display: none or a closed particulars component is out of attain until opened | Ask how your theme hides them, and keep key facts out of hidden sections (SEO, alongside front-end) |
| Structured data (JSON-LD) | No | Chrome leaves manuscript blocks out of its accessibility tree, and Meta says Muse’s browsing delegate says that tree, not the leaf code | Keep it for Google, and put the identical facts in apparent content (SEO) |
| Content rendered by JavaScript | Yes, formerly the browser has rendered it | The browser runs your page’s scripts. The delegate adds none of its own | Keep key facts in the raw HTML too, for crawlers that never run scripts (SEO) |
To see what Muse sees, open a leaf in Chrome DevTools, go to the Elements panel, afterward the Accessibility tab, and toggle “Show accessibility tree.”
I checked the structured-data row on a genuine merchandise leaf from a ample shop on Shopify, study the accessibility tree Chrome 154 built for it. Chrome is built on Chromium, and Meta says Muse runs “a genuine up-to-date Chromium-based browser.” The page’s organized data listed six of the seven sizes in one color as out of stock. A individual sees those six buttons grayed out and traversed out, through styling alone. In the accessibility tree there are seven identical buttons, “Select size eight” to “Select size 14,” alongside nothing marked unavailable. The inventory data existed for Google and for individual eyes, and Chrome’s accessibility tree, the type Meta says the delegate reads, had neither. The tree did capture up following a click: choose size nine and an “Out of stock” note and a “Notify me” clasp appear. An delegate learns what a individual sees at a glance lone by trying all size, one by one.
On the identical page, the shipping eligibility and the free-shipping threshold appeared lone following the page’s scripts ran, so a crawler study the raw HTML would have missed them.
Meta has not stated whether Muse says organized data by another route, and businesses in the connector tier hand their data complete by API without depending on the leaf at all.
Muse Visits Look Like Customers In Your Analytics, And Agent Checkouts Skip Them
| What the customer’s AI did | What your reports show |
|---|---|
| Bought inner an AI app (direct checkout on Meta or Google) | The order, in your platform’s records. The checkout never loads your pages, so Google Analytics and the pixels on your pages don’t run there (Shopify confirms this for its stores on Meta and Google). |
| Sent the individual to your website through a link | A sojourn from that app, if the app passes a referrer. With no referrer and no tag on the link, a direct visit. |
| Browsed and bought on your website (Muse or a dot) | A sojourn from your customer. Neither Meta nor OpenAI publishes a person agent, referrer, or IP range for these browsers, so there’s nothing to display them by. |
Google Analytics 4 added an AI Assistant channel on May 13 that sorts visits from AI assistants specified as ChatGPT, Gemini, and Claude without setup. It plant from the referrer, so it counts lone the second row, and lone whenever the app sends one.
Shopify says its classification of bot sessions in analytics is “designed to be traditional since it’s improved to young female several bots than incorrectly tag genuine customers as bots,” so an delegate meeting is additional apt to district in the individual count. Nobody has published logs showing how GA4 or any bot display labels these sessions.
What The Evidence Can’t Tell You Yet:
- How many group use them. In first October, The Information reported, according to inner Meta data it reviewed, that additional than 3 myriad group immediate Muse at smallest formerly a week. OpenAI has published no numbers for dots. Neither fig counts website visits specifically.
- How many purchases happen. No vendor or third gathering has published how many orders these agents complete, through a browser or a connector.
- How brands get chosen. Neither business documents how its delegate ranks or picks between businesses in an answer.
Muse And Dots Turn Website Structure Into A Budget, Team And Risk Decision
Search engines peruse your website. AI answers citation it. Agents use it. SEO gets a leaf found, and GEO gets it cited in an AI answer. An delegate akin Muse finds the page, says it, afterward keeps going: It picks the size, fills in the form, and pays. GEO stops at the answer, and SEO’s job used to end at the door. Now the leaf SEO got established is additionally the leaf the delegate has to use.
That is additionally why the shortcuts sold for AI visibility young female this visitor. A markdown type of a leaf and an llms.txt document are content prepared for a device to read. They are the AMP leaf all complete again, served to a visitant who came prepared to inspect out. Both depart out the buttons, the forms, and the checkout, which is exactly what an delegate came to use. I wrote concerning this in The Text-Only Version Of Your Website Strips Out The Wrong Layer. Muse says the leaf through its accessibility tree, and neither Meta nor OpenAI says its delegate says either file. For an delegate that came to buy, markdown is pointless.
I called the complete image Machine-First Architecture whenever I wrote concerning it for Search Engine Journal in May. It rests on four parts: identity, structure, content, and interaction. SEO and GEO mostly activity on content, and agents need the another three as well.
For leadership, that method four decisions.
- Team: Put SEO and front-end on the pages anywhere group act: forms, bookings, signups, and checkout. They now build one thing, a leaf a device can peruse and act on. Give SEO a place in the bot-protection assessment too, since that regulation now decides whether a customer’s delegate gets in.
- Risk: Decide on intent whether agents get in. You don’t desire a bot regulation written for scrapers turning distant a customer’s delegate trying to buy from you. Amazon blocked Muse deliberately and stated why. If nobody has decided, a default nobody reviewed is deciding for you.
- Budget: Pay to fix accessibility on the pages anywhere group buy, book, or sign up. WebAIM established that 95.9% of the top myriad homepages had detected WCAG two failures in 2026, so for most websites this was already overdue, before any delegate arrived.
- ROI And Opportunity: Nobody can measure browser-tier delegate income yet, so the case rests on what the activity pays today. That fix serves screen-reader users now, and whenever an delegate compares ten stores, it can lone buy from the ones whose pages it can engage with.
Getting cited is what pays today; I can’t contend that. But OpenAI, Google, and Meta, the companies whose AI answers citation you, are all construction agents that can additionally use your website and equal buy from you. Stop at the citation, and that delegate can citation your website but can’t buy from it.
Key Takeaways For SEO And Marketing Leaders
- Find your tier before anyone asks. Shopify’s default may already have put you in Muse’s connector tier.
- Give SEO a place in the bot-protection review. Your bot rules decide what happens to these agents.
- Audit the pages anywhere group act as an accessibility tree. Put the facts from your organized data in apparent text.
- Report in-app delegate orders from your platform’s command records. Google Analytics misses them.
- Treat this as additional than an SEO project. Front-end, security, and SEO all own part of it.
- Start before the use numbers exist. The connectors, defaults, and browsers are already live.
More Resources:
- Why Google’s New “Google-Agent” Is The Biggest Mindset Shift In SEO History
- Google Tells Developers To Build For AI Agents, Not Just Humans
- Why Agentic AI May Flatten Brand Differentiators
Featured Image: New Africa/Shutterstock
