Cloudflare's new feature forces AI bots to waste resources on an endless maze of false information

Cloudflare has announced a new feature called "AI Labyrinth," which aims to protect against unauthorized data collection by artificial intelligence, Ars Technica reports.

Instead of traditional bot blocking, the system creates a "maze" of AI-generated pages containing neutral scientific information. This allows bots to waste computing resources trying to collect data to train language models without permission.

Cloudflare says that its network handles more than 50 billion requests from such bots every day, which is about 1% of all web traffic. The new feature is available to all of the company's customers, including those on free plans. Importantly, the "maze" uses meta directives that prevent pages from being indexed by search engines, but leave them attractive to bots.

This approach differs from traditional protection methods because blocking bots can alert their operators to detection. Cloudflare believes this method is more effective in protecting data and reducing the risk of unauthorized use of information.