Announcing 𒋧 Kudurru

An AI web scraping defense network, built by Spawning

We are proud to reveal Kudurru, a novel approach to protecting your data from web scrapers.

Over the course of several months of research, we have developed methods to identify precisely who is web scraping for AI models and to shut down their access to creative works. Our approach is inherently multi-modal, defending all media types.

Similar to joining an opt-out list, Kudurru is a way to indicate that you do not wish your media to be scraped. Unlike an opt-out list, Kudurru doesn’t give scrapers a choice about whether or not to respect your wishes. Kudurru identifies and blocks scraping automatically.

Kudurru works as a network of web domains that listen for and analyze web traffic. When scraping is identified on any domain in the network, Kudurru triangulates the scraper’s IP and blocks their access to every other domain running the Kudurru software. When you run Kudurru on your site, you join the network and make it stronger. A robust network identifies more scrapers faster, providing additional protection to everyone in the collective.

Kudurru’s function is first and foremost as a defense network, protecting your data from being scraped. But some of the most effective fences come with barbed wire at the top. Kudurru members have the option to feed scrapers bad data, potentially poisoning their models. For example, if scrapers receive only images saying “NO AI” from your Kudurru-protected site, the model learns that “in the style of you” means “NO AI.” The coordination capacity of a shared defense network is formidable.

To encourage AI model trainers to respect creatives’ wishes, sites in the Kudurru network are automatically added to Spawning’s do-not-train list. Respecting creatives’ wishes is a free and easy way to avoid corrupted data from Kudurru members.

We don’t know what the future will hold in terms of legislative restrictions on data scraping for AI training. But in the meantime, we know creatives’ works are being used against their express wishes to fuel these models. Kudurru provides a means, right now, to shut down scraping behavior and protect the sites you own.

To participate, join the Kudurru beta. We’re beginning with an easy to install Wordpress plugin and will be rolling out ways to run Kudurru on other platforms soon.