Cloudflare introduces differentiated AI crawler control by purpose (heise.de)

• Cloudflare introduces three AI crawler categories: Search, Agent, Training
• By Sept 2026, training/agent crawlers blocked by default on ad-supported new domains
• Enterprise BotBase database classifies bots and usage levels

"Cloudflare expands AI crawler controls: website operators can now distinguish and block access by purpose (search, agent, training). From September 15, 2026, new domains will block training and agent crawlers by default on ad-supported sites, while allowing search engine crawlers. Cloudflare also introduces BotBase, a database of verified bots with classifications and usage levels. The features are available to all customers, including free tier."