On a site drawing a million page views, AI traffic can account for 15% of all requests. And generate zero ad revenue. The servers still run. The bandwidth still moves. The CPUs still spin. Yet the ad network sees none of it. AI agents skip ad slots, ignore tracking scripts and never trigger the click that converts a page load into income.
That gap has a number. Apply a $1 CPM to the AI-driven portion of a million-page-view site and the monthly figure sits around $150. Invisible under standard analytics, but real. Multiply it across a mid-sized publisher running several properties and the uncaptured value becomes harder to ignore.
The problem runs deeper than missed ad impressions. Subscriptions and paywalls assume a reader who returns, pays monthly and wants the full article. AI agents want none of that. They pull a paragraph, a product specification or a structured data snippet and leave. Agents have no interest in a monthly pass. Nor will they fill in a signup form. Traditional revenue models did not account for this kind of visitor, and they capture no value from it.
Publishers who try robots.txt rules or rate limits find the same result. Bots slow down or route around the restrictions. Cached copies surface elsewhere and the original source loses attribution without recovering any revenue. Licensing deals sound more promising in theory, but in practice they favour the largest platforms. Smaller publishers rarely receive offers, and managing thousands of individual agreements at scale has no infrastructure behind it.
PayLayer approaches the problem differently. The WordPress plugin intercepts requests from known AI agents and responds with a 402 Payment Required status code. That HTTP response code has existed for years but rarely found practical use until now. Alongside that status, the response includes payment details in x402 headers. The agent pays the stated micro-fee, the resource streams through, and human visitors never see any of it. Their browsing, their caching and their experience stay entirely untouched.
The pricing sits at fractions of a cent per request. An article summary returned as JSON costs $0.0005. A full HTML body costs $0.002. A live product feed with real-time inventory data costs $0.01 per request. A dedicated route lists all available prices so agents can check costs before calling any endpoint. Think of it as a machine-readable menu that removes friction from the transaction. Publishers set the rates and can adjust them per endpoint.
Meanwhile, WooCommerce stores get additional options. PayLayer allows bots to browse product catalogues, add items to carts and complete checkouts through paid API calls. Each step carries a fee. Every completed order carries a notation marking its AI origin. That suits businesses running automated restocking or procurement workflows that need clean provenance on machine-generated orders.
In turn, the use cases extend across several publisher types. A site with heavy article traffic can expose clean JSON endpoints for AI summarisers. RAG pipelines — the retrieval systems behind AI assistants — pay per response instead of scraping the same content for free. A product catalogue can set lower fees for cached data and higher rates for live stock checks. A publisher with a media library can charge per megabyte of image or transcript delivery. Signed URLs handle distribution through a content delivery network.
Beyond article and product data, PayLayer includes a citation endpoint model. Verifiable quote snippets with cryptographic hashes sell at small amounts per request. AI agents get authoritative sourced text, the publisher gets revenue and attribution stays intact.
The broader argument behind PayLayer is a reframing rather than a technical fix alone. AI agents consuming content are not a nuisance to block. They signal real demand, rely on accurate information and place genuine load on infrastructure. Treating them as paying customers rather than resource drains changes the economics of publishing a website. A bot paying $0.001 per article request across a thousand daily visits generates predictable income. Human visitors who load three pages and ignore four ads do not.
No named founder appears in PayLayer’s public materials, which publishers evaluating the plugin should note. The technology draws on the x402 standard — a real HTTP protocol with genuine industry support. Its pricing model is specific enough to test against actual traffic data.
