お知らせ • Jul 03
Cloudflare, Inc. Announces New Classifications, Enhanced Analytics, and Industry-Defining Commercial Partnerships
Cloudflare, Inc. announced new classifications, enhanced analytics, and industry-defining commercial partnerships that bring together site owners and transparent AI companies so that the agentic Internet can flourish. Cloudflare's new tools and integrations help site owners and AI companies optimize for discoverability, efficiency, and monetization. Cloudflare is testing new default classifications, delivering deeper insights to customers, making AI search faster, and ensuring creators are compensated when their content powers an answer. Cloudflare will engage with the ecosystem, listen to feedback, and run tests to finalize new defaults and classifications with a deadline of September 15, 2026. For new customers and for new sites for existing customers, on September 15, 2026, the defaults will be set to allow for search but block training and agent use for pages with ads. Mixed crawlers that do not give site owners the ability to choose between search, agent use, and training, will be blocked on all pages with ads. At any time these settings can be changed by customers in their dashboard. On September 15, 2026 these changes will also be made for all existing free customers that have not changed their settings by September 15, 2026 in their dashboard. Cloudflare is introducing the new Attribution Business Insights dashboard, stripping away the data asymmetry that has long left content owners negotiating in the dark. For the first time, business stakeholders, not just security and IT teams, can clearly see how AI bots consume their content and track how much human traffic specific AI companies actually send back. Cloudflare aims to provide the definitive measurement and optimization layer for the AI era. Cloudflare is working on two fronts that help creators and AI companies alike: making AI search more efficient, and paying creators automatically when their work is used. Cloudflare data suggests that over 50% of crawl traffic from AI crawlers is spent re-fetching unchanged pages. Because Cloudflare can see what has actually changed, it can signal when a page is worth re-fetching and when it isn't — cutting needless crawls in the process, while delivering better results. Cloudflare is testing these signals with leading AI companies to measure how much they improve freshness and reduce wasted crawling, and plans to make them broadly available later this year. Cloudflare launched the first version of Pay Per Crawl so publishers could charge AI companies to crawl their content. Cloudflare is evolving Pay Per Crawl into Pay Per Use, so publishers are paid when their content actually creates value, not just when it's fetched. The company is working with top partners, including Ceramic.ai and You.com to turn that principle into a working market. With Ceramic.ai, when a publisher opts in, they are paid every time their content appears in Ceramic's AI search results, and they get back the queries, citations, and ranking back. Other AI companies can bring the model that fits how they work — You.com, for instance, will let an agent pay on demand for a specific piece of premium content it needs, while Cloudflare provides the common layer beneath them all. Cloudflare introduced AI Crawl Control, which set a permission-based model for AI crawling. Cloudflare partnered with platforms like beehiiv, providing independent creators and major media outlets alike with granular dashboards to easily track, block, or opt into specific AI models. Cloudflare also spearheaded Web Bot Auth, an open framework that allows AI bots and autonomous shopping agents to verify their identity and legitimacy to publishers and merchants. Cloudflare has also collaborated with leading payment providers to lay down the underlying payment rails for the emerging agentic commerce economy. Building directly upon Web Bot Auth, Cloudflare is helping them ensure AI shopping agents can transact safely at millions of merchants globally without exposing raw credentials.