Cloudflare found in June that automated bots now make up more than half of all internet traffic. The figure is rising as AI companies scrape the web and agents attempt to perform tasks on behalf of users. Cloudflare positions itself between websites and these agents, offering owners control to block all AI tools, permit them, or restrict access to those willing to pay.
Who is Matthew Prince
Matthew Prince is the CEO and co-founder of Cloudflare. He joined the Decoder podcast two and a half years ago during a period described as a pivot point for the internet. The situation has since become more complex due to the rise of AI. Prince previously discussed free speech and 8chan with the show. He returned for a new conversation about business models, bot traffic, and the future of the web.
The shift in traffic
Prince stated he usually gives a short answer at cocktail parties to end conversations. When speaking with interest, however, he describes Cloudflare as an effort to rebuild the internet as it should have been constructed from the start. The company operates a global network spanning over 350 cities with thousands of data centers. Its core functions include securing applications and content, ensuring safety for users online, and enabling applications to scale to a global audience.
The business model of the internet has changed dramatically due to AI and bot traffic. For at least 30 years, advertising has driven web growth. Google is credited with defining this model for 28 years through search and its advertising ecosystem, including DoubleClick and Google Analytics.
Prince attended Web Summit last November, where he was asked when non-human traffic would exceed human traffic. Data collected by Cloudflare indicated this would happen in the second half of 2027. By March 2026 at South by Southwest, the prediction had moved to the first half of 2027. A few months later in May, the team informed Prince that automated traffic had already surpassed human traffic online.
Prince noted that extrapolating current trends suggests automated traffic could reach 1,000 times human traffic in five years. This assumes human traffic remains stable while automated activity explodes. He acknowledged that previous predictions regarding this specific metric have been incorrect.
Infrastructure costs
Managing 1,000 times more traffic requires significant infrastructure investment in bandwidth and servers. The traditional advertising model does not work for bots, which do not click on ads or respond to traditional branding. Prince is focused on determining a new payment scheme to power this insatiable demand.
When asked if revenue and profits will exist, Prince confirmed there is certainly a lot of revenue. He noted that Anthropic added $10 billion in revenue within a single month. AI companies require three main resources: talent, silicon, and data.
Scarcity and supply
Talent is currently scarce, though every university is expanding its AI departments and labour markets are efficient enough to train new workers. Silicon availability is also constrained. Nvidia currently leads in chip production, but competitors like AMD, Qualcomm, and hyperscalers are building their own silicon. Supply shortages are expected to ease over time.
Data is the third requirement. Prince indicated that the availability of data is moving in a different direction compared to talent and hardware. Historically, the internet has generated vast amounts of data, but the nature of this generation is changing as the web is increasingly populated by automated agents rather than humans.
Internal changes
AI is also restructuring operations inside Cloudflare. Earlier this year, the company laid off more than 1,000 employees, representing 20 percent of the workforce. Prince wrote an op-ed for the Wall Street Journal titled “How I choose which Cloudflare employees to replace with AI.” This decision followed a period of significant growth and subsequent adjustment in the company’s structure.
Prince argues that Cloudflare is helping shape the future business model of the internet to ensure it remains healthy. The goal is to transition from a model driven by human attention and advertising to one that accommodates the massive volume of automated traffic currently defining the web.
