Claude Fable 5: Testing the First Publicly Available Mythos-Class Model: Can They Find Your Business?

A glowing magnifying glass scanning miniature business storefronts connected by data lines and location pins

⏱ 8 min read · Last updated 2026-06-11

This week a new, more capable class of frontier AI models reached general availability — and for small businesses, the headline isn’t the benchmark scores. It’s what these systems now do. They no longer just answer a question and stop. They browse the open web, read charts and screenshots, hold their focus across millions of tokens, and increasingly take multi-step actions on a person’s behalf. The practical question for every operator is suddenly very concrete: when one of these agents goes looking for a business like yours, can it find you, understand what you sell, and reach you?

Why It Matters

For two decades the discovery game was simple: rank on a page of blue links, earn the click, convert the visitor. That funnel is being rewired. A growing share of customer research now happens inside an AI assistant that reads, summarizes, and decides — often without the user ever visiting your site.

The scale of the shift is no longer speculative. Gartner has forecast that traditional search engine volume will drop 25% by 2026 as people route queries through chatbots and virtual agents instead. And when an AI summary does appear in a results page, people click through far less: Pew Research found that users clicked a traditional link on just 8% of pages with an AI summary, versus 15% without one. The intermediary is becoming the destination.

For an SMB, that means your visibility increasingly depends on whether a machine can parse you, not just whether a human can scroll to you. If the agent can’t confidently identify your hours, your service area, your offer, and a working way to contact you, it routes the customer to a competitor it can read.

What’s New / How It Works

The leap in this latest model generation is agentic competence. These systems can take a goal — “find a licensed HVAC company near me that does emergency weekend calls and book a slot” — and execute it as a chain of steps rather than a single lookup. Three capabilities matter most for your discoverability:

Open-web browsing. The agent visits real pages, reads your listing, and cross-references directories instead of relying only on a frozen training snapshot.
Vision and layout parsing. Newer models read screenshots, scanned menus, charts, and image-based contact blocks — but text it can copy and verify still wins over a phone number trapped inside a graphic.
Long-horizon action. The agent holds context across a long task: comparing five vendors, weighing reviews, and shortlisting the ones with clean, consistent, machine-readable details.

The mechanics of being chosen come down to legibility. An agent rewards a business whose name, address, and phone number match everywhere it looks, whose listing is claimed and complete, and whose contact path actually resolves. Inconsistency reads as risk, and the agent quietly drops you from the shortlist.

When an AI agent shops on your customer’s behalf, the business it can read and reach wins — not the one that ranks best for humans.

The Numbers

The behavioral data behind this shift is what makes it urgent:

25% projected drop in traditional search engine volume by 2026 as queries move to AI agents, per Gartner.
8% vs. 15% — the click-through rate on pages with an AI summary versus pages without, according to Pew Research Center.
1% — the share of users who clicked a link inside the AI summary itself, per the same Pew analysis, showing how rarely the AI hands the click back.
26% vs. 16% — how often users simply ended their browsing session after a visit that included an AI summary, compared with visits without one (Pew).

The funnel didn’t disappear — it moved inside the assistant. The businesses that stay visible are the ones a machine can identify, trust, and contact without friction.

What Comes Next

The next stage is transactional. Agents are moving from recommending a business to contacting one — filling a quote form, sending an inquiry, even initiating a booking. That turns “can the AI find me?” into “can the AI actually reach me and get a response?” A contact form that breaks, an inbox no one monitors, or a chat widget that demands a human-only CAPTCHA becomes a dead end for an agent acting on a real customer’s intent.

Expect rollouts to stay phased and capabilities to keep climbing, with providers layering in safety controls and data-handling rules along the way. None of that changes the operator’s job. The durable work is making your business unambiguous to a machine: claimed listings, consistent details across every directory, structured data on your site, and a contact path that responds. That groundwork pays off regardless of which model is on top this quarter.

What This Means for You

You don’t need to predict the next model release. You need to be readable and reachable today. Start by checking whether AI agents can actually contact you — run an AI-contactability check to see if your forms, phone, and contact data resolve cleanly for an automated visitor, not just a human one. Then make sure you’re even in the consideration set: claim and complete your business listing so your name, address, and hours are consistent everywhere an agent looks, and tighten the directory citations behind your business listings so nothing contradicts.

This is the same pattern we’ve traced as search fragments across AI surfaces. If you haven’t seen them, our breakdowns of how AI Mode and AI Overviews drive opposite behavior and the finding that 88% of users accept AI Mode results as-is explain why being the cited answer now matters more than ranking. And because agents increasingly judge a brand by whether it looks active and consistent, keep your social profiles current with Feedsta, an AI social media manager that creates, schedules, and analyzes posts across platforms so your brand stays visible wherever customers — and their assistants — go looking.

The Bigger Picture

The arrival of more capable AI agents isn’t a story about benchmarks — it’s a story about who gets chosen when a machine does the choosing. The advantage no longer goes to the business that games a ranking; it goes to the one that is legible, consistent, and contactable end to end. That’s squarely within your control, and the operators who lock it down now will be the default answer while everyone else is still waiting to be found.

Frequently Asked Questions

What does it mean for an AI agent to "find" my business?

It means an AI assistant, while completing a task for a user, can locate your business, correctly interpret what you offer, and identify a working way to reach you. Unlike a human who scans a results page, an agent reads structured data, directory listings, and your site’s contact details programmatically. If your name, address, phone, and hours are consistent and machine-readable across the web, the agent can confidently recommend or contact you. If those details conflict or are buried in images, the agent treats you as unreliable and routes the customer to a competitor it can parse more easily.

Why are AI agents reducing clicks to my website?

Because the assistant often answers the user inside the chat or summary instead of sending them to your page. Pew Research found people clicked a traditional link on only 8% of pages that showed an AI summary, versus 15% without one, and just 1% clicked a link inside the summary itself. The customer gets their answer from the AI, which may have read your content without ever passing along the visit. That makes being the source the AI cites and trusts more valuable than simply ranking, since the click you used to count on may never arrive.

How do I make my business readable to AI agents?

Focus on consistency and structure. Claim your listings on major platforms and keep your name, address, and phone number identical everywhere. Add structured data (schema markup) to your website so machines can extract your hours, services, and location without guessing. Keep contact information as real, copyable text rather than locked inside an image. Audit directory citations and fix anything that contradicts your site. The goal is to remove ambiguity: an agent rewards the business whose details line up across every source it checks and drops the one whose information conflicts.

What is AI-contactability and why does it matter now?

AI-contactability is whether an automated agent can actually reach your business, not just find it. As agents move from recommending businesses to contacting them, filling forms, sending inquiries, and starting bookings, a broken contact form or an unmonitored inbox becomes a dead end. If the agent can’t complete the contact step, it abandons you mid-task even if you were the best match. Testing your contact paths for automated visitors, not only human ones, is becoming as important as ranking was a decade ago, because the agent represents a real customer with real intent.

Is traditional SEO dead because of AI agents?

No, but its role is changing. Clean content, fast pages, and authoritative links still help machines understand and trust you. What’s shifting is the payoff: instead of optimizing only for a click from a ranking, you’re optimizing to be the answer an AI selects and cites. Gartner projects traditional search volume will fall 25% by 2026 as queries move to AI agents, so leaning entirely on blue-link rankings is risky. The smarter play is to keep solid SEO fundamentals while making your business explicitly legible and contactable for AI systems.

Do I need to worry about which specific AI model is most powerful?

Not really. Model leaderboards change every quarter, and chasing the top performer is a distraction for most operators. What persists across every model is the requirement that your business be unambiguous and reachable: claimed listings, consistent details, structured data, and working contact paths. That groundwork benefits you no matter which assistant a customer uses or which model is leading benchmarks this month. Build for legibility and contactability once, and you stay discoverable as the underlying technology keeps advancing.

How can a small business start preparing today?

Begin with an audit. Check that your business is listed and complete on the major directories and that the details match your website exactly. Add or verify structured data so machines can read your hours, location, and services. Test your contact forms, phone, and chat to confirm they resolve cleanly. Review whether your most important pages clearly state who you serve and what you offer in plain text. Finally, keep your profiles active and consistent so an agent sees a current, trustworthy brand rather than a stale or contradictory one.