Cloudflare CEO Warns AI Could Destroy the Internet’s Open Future

Cloudflare CEO Matthew Prince has issued a stark warning about the future of the internet, describing a scenario where artificial intelligence could fundamentally undermine the open, accessible nature of online information. Speaking on WIRED’s Big Interview Podcast, Prince argued that the era of search engines as the primary gateway to the web is over. Today’s AI-powered “answer engines” deliver synthesized summaries directly at the top of search results, replacing the traditional list of links that once guided users to original content. Prince sees three possible futures for the internet. The first—where all human-created content vanishes and is replaced by AI-generated material—is unlikely, he says, because AI still depends on human-written content to learn and operate. The second, which he calls “frighteningly likely,” draws a chilling parallel to the historical model of the Medici family in Renaissance Florence. In this “Black Mirror” scenario, a small number of AI companies—such as OpenAI, Anthropic, and Perplexity—would become the dominant patrons of content creators, hiring journalists, researchers, and writers to produce material for their platforms. This would concentrate control over knowledge in the hands of a few powerful corporations, potentially shaping information to align with their own ideologies. Prince points to recent evidence of such bias, citing a study that found xAI’s Grok chatbot often reflected Elon Musk’s personal views when answering sensitive questions. He warns that this could lead to a fragmented information landscape, with distinct AI-driven “worldviews”—a conservative one, a liberal one, a Chinese one, an Indian one—each catering to different audiences. This would erode the internet’s original promise as a global, neutral space for free and open exchange of ideas. The third, more hopeful scenario, is one where content creators regain control by establishing scarcity. Prince believes AI companies will eventually need to pay for access to human-generated content, much like Netflix licenses shows. This shift is already underway. Cloudflare has introduced tools that allow website owners to block AI crawlers unless they pay for access, a move aimed at forcing AI firms to compensate creators. Major publishers like the Associated Press and Condé Nast are already using these tools. This summer, Penske Media Corporation, owner of Rolling Stone and The Hollywood Reporter, sued Google over its AI Overview feature, accusing it of scraping content without permission. Prince stresses that this outcome is essential for Cloudflare’s survival and for the internet’s future. “If the internet stops existing, what’s left for Cloudflare to do?” he asked. Yet even this hopeful path raises concerns. If news organizations become content suppliers for AI systems, will the focus shift from public service journalism to maximizing data value for AI training? The challenge now is not just about compensation, but about preserving the integrity and diversity of information in an age where AI shapes how we see the world.