Policy

Cloudflare CEO: Bot Traffic Now Exceeds Human Web Use by 60%

Matthew Prince warns automated requests could outnumber people 1,000-to-1 within five years as AI crawlers consume content without compensating creators.

Omega Editorial· September 18, 2026· 3 min read

Automated bot and AI agent traffic overtook human visitors on the web in May 2025, reaching nearly 60% of all website requests—months ahead of Cloudflare CEO Matthew Prince's original projection for late 2026.

At the current trajectory, Prince estimates automated traffic could exceed human activity by a factor of 1,000 within five years. The shift poses fundamental questions about how content creators get compensated when AI crawlers ingest their work to train models that may never return readers or revenue to the original source.

"Companies need to get paid by the AI companies for what is the fuel that runs these AI systems," Prince told Fortune from his office at One World Trade Center in New York.

New controls for publishers

Cloudflare, which manages traffic for more than one-fifth of the web and counts 80% of major AI companies as customers, announced a "Disallow AI Training" setting this week. The tool allows websites to remain indexed for search while blocking crawlers from using their content for model training. Apple, Google, and Microsoft have agreed to honor the mechanism.

The announcement represents one of the first practical tools publishers have to distinguish between crawlers that index content for search discovery and those that extract it for AI training. The timing coincides with ongoing copyright litigation from Sony, Warner Music, news organizations, and dictionary publishers against AI companies over intellectual property claims.

Why it matters

The economics of content creation break down when AI systems consume vast amounts of material without attribution or payment. Publishers lose both direct revenue and the indirect benefits of audience relationships. As Prince noted, creators often value recognition as much as compensation—making anonymized extraction particularly damaging. Without sustainable models that reward the labor and cost of producing original work, the infrastructure that generates reliable information becomes economically unviable. Prince's warning extends beyond media: "If we don't have a way of paying content creators, of funding the infrastructure buildout, of making everything more efficient, I just don't know what the internet looks like in the future."

Credibility over popularity

Prince argues the next phase of digital media will reward credibility rather than popularity—a shift driven by AI's ability to generate millions of derivative articles. "Bots can spit out millions of articles that riff off the news," he said. "Someone has to report what's really going on."

He suggested the industry needs recognition systems comparable to Academy Awards or Nobel Prizes to provide proper attribution for knowledge creation. His vision anticipates a future with 500,000 AI companies rather than five dominant players, requiring robust mechanisms for both compensation and credit.

These details were first reported by Fortune's CEO Daily newsletter.

#cloudflare#web traffic#ai crawlers#content monetization#bot traffic#matthew prince

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Policy

Policy· 3 min read

Robot Tax Proposals Gain Traction as Automation Displaces Workers

Levies on automation could fund social safety nets, but implementation challenges and political resistance remain formidable obstacles.

Via Automation Watch · Sep 18, 2026
Policy· 4 min read

Congress Needs In-House AI Expertise, Not Industry Lobbyists

Without rebuilding technical capacity gutted in 1995, lawmakers risk letting frontier AI companies write their own regulatory playbook.

Via AI Watch · Sep 18, 2026
Policy· 3 min read

AI Safety Evaluator Debate Intensifies Amid Regulatory Vacuum

White House pushes industry self-policing while questions mount over who should independently assess frontier AI systems.

Via AI Watch · Sep 18, 2026