Cloudflare acquires Human Native to structure licensed AI data

Cloudflare has acquired Human Native, a UK-based AI data marketplace focused on turning unstructured multimedia content into searchable, licensable data. The Human Native team has spent the past several years helping publishers and developers turn raw content into structured assets that can be licensed and valued — treating data not as something to be scraped, but as an asset class requiring transparency and control.

The acquisition points to a practical benefit beyond licensing convenience: better data quality. One of Human Native’s customers, a UK video AI company, reportedly discarded its existing training data entirely after getting superior results from fully licensed, reputably sourced content via Human Native. That customer now trains exclusively on such material.

Why the current web economy is under strain

The open Internet has run for three decades on a simple exchange: creators publish, aggregators send traffic, and creators monetize that traffic via ads, subscriptions, or direct support. That loop is now under real pressure. Crawl-to-referral ratios have skyrocketed — tens of thousands of AI and bot crawls for every real human visitor — and it is often unclear how multi-purpose crawlers use the content they access.

Publishers, news outlets, financial firms, tech companies, and other creators share a common requirement: they want a say in how AI systems use their content. Cloudflare’s existing work on AI Crawl Control and Pay Per Crawl is built on that premise. Those tools, along with AI Search, let content owners decide whether they want maximum AI visibility — appearing in every training set and new search — or restricted access with direct compensation. The principle is simply that the content owner chooses.

Structured data replaces blind crawling

With Human Native on board, Cloudflare is accelerating work on helping customers make content easily accessible to AI bots and agents alongside human audiences. Crawling the open web is costly in engineering and compute, and offers no quality guarantees: crawled indexes routinely contain duplicates, spam, and illegal material, leaving developers with messy, unstructured data.

Cloudflare’s recently announced AI Index is designed to address that. Rather than sending crawlers repeatedly across the open web, the AI Index uses a pub/sub model: participating sites expose structured updates whenever content changes, and AI developers subscribe to receive those updates in real time. This gives developers a cleaner alternative to traditional crawling and opens new business-model options for content creators.

Machine-to-machine payments as a foundation

Supporting those new models requires payment infrastructure that works between automated systems, not just humans. Cloudflare is building the x402 Foundation with Coinbase to enable machine-to-machine transactions for digital resources. Traditional web payments assume human intent: browse, add to cart, enter card details, click pay. Direct transactions between automated systems need protocols of their own.

BLOG-3111 1

The combination of Human Native’s data marketplace expertise with Cloudflare’s crawling control, indexing, and payment work is aimed at establishing the basis for these new economic models. The stated goal is an Internet that remains open, fair, and independently sustainable in the age of generative AI — with better AI built on better data, and creators retaining control, compensation, and credit.