Cloudflare CEO warns Artificial Intelligence crawlers threaten the web’s business model

Cloudflare´s CEO calls attention to the existential risk posed by Artificial Intelligence crawlers and summaries, which divert traffic away from websites and imperil digital publishing revenues.

Cloudflare CEO Matthew Prince has renewed warnings about the growing impact of generative artificial intelligence crawlers and automated summaries on the viability of the internet’s traditional business model. Speaking at an Axios event in Cannes, Prince highlighted an accelerated decline in web traffic resulting from chatbots and search engines that scrape and condense web content, reducing the number of genuine human visitors to publishers’ sites. He noted a worsening trend even compared to six months earlier, with decreasing ratios of real visits compared to automated crawls from both established search engines and emerging artificial intelligence models.

Prince provided stark metrics illustrating the scale of the issue: whereas previously, a Google crawl would yield one visitor for every six scans, that ratio has dropped to 18 to 1. The disparity is even more pronounced with companies like OpenAI and Anthropic, where the rates have plummeted to 1,500 to 1 and roughly 60,000 to 1, respectively. The root cause, according to Prince, is the increasing dominance of instant answers—users rely on chatbot summaries and search engine overviews rather than clicking through to original content, effectively bypassing ad-driven revenue models that sustain much of the web’s publishing ecosystem.

While tech giants such as Google and large language model developers have asserted that providing citations will in fact bolster traffic to the sources of Artificial Intelligence summaries, Prince argues that such measures have had negligible impact: most users simply trust the summarized responses and rarely follow source links. This not only deprives publishers of visitors and advertising income but also carries risks due to Artificial Intelligence´s propensity for sometimes generating misleading information. In an effort to defend web publishers, Cloudflare launched a tool called AI Labyrinth, which uses generative techniques to bait non-compliant Artificial Intelligence crawlers into wasting computational resources on algorithmically generated mazes of irrelevant content, mitigating the drain on legitimate sites. Despite the challenge of opposing major Artificial Intelligence ecosystem players, Prince affirmed Cloudflare’s ongoing commitment to protecting its clients and maintaining a fair, open web.

74

Impact Score

Introducing Mistral 3: open artificial intelligence models

Mistral 3 is a family of open, multimodal and multilingual Artificial Intelligence models that includes three Ministral edge models and a sparse Mistral Large 3 trained with 41B active and 675B total parameters, released under the Apache 2.0 license.

NVIDIA and Mistral Artificial Intelligence partner to accelerate new family of open models

NVIDIA and Mistral Artificial Intelligence announced a partnership to optimize the Mistral 3 family of open-source multilingual, multimodal models across NVIDIA supercomputing and edge platforms. The collaboration highlights Mistral Large 3, a mixture-of-experts model designed to improve efficiency and accuracy for enterprise artificial intelligence deployments starting Tuesday, Dec. 2.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.