Cloudflare has introduced a "Disallow AI Training" setting to resolve the conflict faced by website owners using mixed-use crawlers, such as Applebot, Bingbot, and Googlebot, which serve both search indexing and AI model training. Previously, blocking these crawlers to prevent AI training inadvertently removed a site from search results. The new granular control uses Bot Preference Sync to publish a "Disallow" directive in robots.txt, allowing "Accountable" operators—including Apple, Google, and Microsoft—to respect training opt-outs while maintaining search discoverability. Cloudflare classifies bot behavior into three distinct categories: Search, Training, and Agent. While training-only crawlers from companies like OpenAI and Anthropic can be blocked without impacting search, mixed-use crawlers require this specific new setting. The update also clarifies that "Block" and "Block on pages with ads" settings will now apply to all training crawlers, including mixed-use ones, as the new AI training setting handles the search preservation. Cloudflare's data highlights the necessity of these controls, noting that while 17% of sites already use some mechanism to block training, less than 1% block search bots. Moving forward, Cloudflare aims to provide even more granular control over AI summaries by early next year.