The rapid rise of artificial intelligence answer engines has introduced a fundamental economic friction to the modern web. When an AI system reads a publisher’s web page and summarizes the information directly for a reader, the traditional website visit—and the digital advertising revenue that accompanies it—often disappears entirely.
To bridge this gap, web infrastructure and security giant Cloudflare is rolling out a beta program known as "Pay Per Use." Designed to establish an economic foundation for the emerging "agentic web," the system provides a scalable mechanism for publishers to state a clear condition to AI developers: yes, you can use our work, but only if you pay for it.
For most digital publishers, negotiating individualized licensing agreements with the countless entities deploying AI products is an impossible administrative burden. Conversely, AI companies cannot practically execute contracts with millions of independent websites. The Pay Per Use model aims to solve this scaling problem by acting as a streamlined commercial layer, connecting willing content creators with verified buyers through unified infrastructure.
The initiative builds upon Cloudflare’s earlier introduction of "Pay Per Crawl," which debuted in 2025. While Pay Per Crawl charges AI entities for basic access and indexing, the new Pay Per Use model shifts the financial focus to downstream utility. Because AI search engines and agents frequently crawl far more pages than they ultimately incorporate into answers or reports, charging solely for every crawl can force buyers to pay before knowing the true value of the data. By tying costs to actual utilization, Cloudflare hopes to attract more commercial buyers while securing sustainable revenue for publishers.
How Pay Per Use Works
The mechanics of the beta program are structured to minimize friction for both content owners and AI developers. Under the framework, an AI company establishes a program with Cloudflare by defining its crawling identity, specifying the exact type of downstream use it intends to pay for, setting a corresponding price, and connecting a funded payment account.
Publishers can review these incoming offers directly within the Cloudflare dashboard under a dedicated monetization section. The interface displays the identity of the AI company, the specific use case covered by the offer, and the proposed price per interaction. Content owners retain full autonomy to evaluate these terms, accept or decline proposals, and withdraw from agreements if an arrangement ceases to be commercially viable.
Crucially, the system does not require publishers to alter their site architecture or implement complex technical integrations for each individual AI partner. Once a publisher accepts an offer, the participating AI company tracks its interactions using verified bots and reports its utilization via a standardized application programming interface.
The scope of what constitutes a paid use can vary depending on the product. For instance, a conversational search engine might trigger a payment obligation whenever it displays a direct excerpt from an enrolled page to an end user. Alternatively, a specialized shopping agent might compensate a publisher when an online product review influences a consumer recommendation, even if the user never directly navigates to the underlying review page. Because a single piece of journalism, research, or commentary can deliver value across entirely different products, publishers have the flexibility to accept varied payment offers tailored to specific use cases.
Reporting and Settlement
To keep administrative overhead to a minimum, the reporting process relies on self-reported data verified against publisher enrollments. Participating AI companies fetch the list of domains that have formally accepted their commercial offers and subsequently submit usage logs.
These reports are transmitted as lightweight data records detailing the timestamp of the interaction, the specific source URL, and a unique event identifier. Cloudflare then aggregates these reported occurrences, handles the billing for the purchasing entity, and distributes payments to publishers on a monthly basis through their connected payment accounts.
This centralized architecture frees buyers from managing thousands of fragmented bilateral contracts and technical integrations, while offering publishers a single dashboard to monitor earnings, track utilization trends across different domains, and evaluate the financial return of their digital catalog.
Connecting Discovery, Utility, and Revenue
Beyond straightforward financial settlement, Cloudflare is positioning the Pay Per Use initiative as part of a broader analytical toolkit for digital publishers navigating the AI transition. The company’s existing business insights tools already provide visibility into which automated crawlers visit a site and the volume of data they consume. Pay Per Use extends this oversight into the post-crawl phase, illuminating whether indexed content was actively utilized, how frequently, and the exact revenue generated by those interactions.
Cloudflare is also working alongside AI developers to provide publishers with richer contextual metadata regarding each paid use. This may include anonymized insights such as the topical keywords that led to a citation, the broader theme of a user request, or the specific category of agentic product that powered the interaction, all while maintaining strict privacy standards that exclude personal data.
When combined with existing tools like Answer Engine Optimization insights—which track how virtual assistants respond to inquiries about a publisher’s domain—the platform attempts to connect three critical pillars of modern publishing: how content is discovered, how it is deployed by autonomous agents, and what financial return it commands. This data equips organizations to make informed commercial decisions regarding which types of content generate sustainable returns, alongside editorial choices concerning what subjects to prioritize, update, or optimize for agentic discovery.
Building an Economic Layer for the Agentic Web
During the initial beta phase, Cloudflare is collaborating directly with participating buyers and publishers to refine the system at scale. The core objective is to ensure the framework delivers mutually beneficial outcomes: providing AI companies with reliable access to rapidly updating, high-value content such as breaking news, original research, and specialized trade publications at a sustainable price, while guaranteeing that publishers receive trustworthy reporting and timely compensation.
As the program evolves, the platform aims to streamline the onboarding process for new buyers, enabling them to introduce novel products and payment models without duplicating administrative workflows. For publishers, this long-term vision promises a centralized marketplace where they can evaluate offers, negotiate pricing dynamically, and attribute varying economic values to different types of content—such as distinguishing the high real-time value of breaking reporting from the historical value of archival material.
Ultimately, the initiative seeks to move the digital ecosystem beyond a rigid binary choice between blocking automated agents entirely or surrendering valuable intellectual property without compensation. By establishing clear commercial terms and transparent records of utilization, Pay Per Use offers a viable path for content creators to sustain their operations within the rapidly expanding landscape of AI-driven tools.
Leave a Reply