Introduction: A New Digital Enclosure Movement
The rapid ascent of generative artificial intelligence has fundamentally altered the landscape of the internet, transforming the vast, decentralized repository of human knowledge into a concentrated asset class. According to former Google engineer and high-profile whistleblower Zach Vorhies, this transition is not the result of organic market evolution, but rather the consequence of a coordinated effort between Silicon Valley’s leading AI firms—specifically OpenAI and Anthropic—and the United States government.
In a recent interview, Vorhies, who gained international notoriety in 2019 for leaking nearly 1,000 pages of internal Google documents, characterized these entities as "data cartels." He argues that these firms have systematically scraped the entirety of the internet’s accessible human knowledge, essentially "closing the door" behind them. By doing so, they have transformed public information into a gated utility, leaving researchers, startups, and the general public reliant on a meter-based subscription model for data that was originally harvested for free.
The Chronology of Information Capture
To understand the gravity of Vorhies’ allegations, one must look at the timeline of the "Great Scrape." In the decade leading up to the 2022 explosion of Large Language Models (LLMs), the internet functioned as a chaotic, semi-open ecosystem. During this window, tech giants operated with minimal regulatory oversight, vacuuming up public discourse, academic research, and creative content.
- 2015–2019 (The Ingestion Phase): Major AI firms began building proprietary datasets by crawling the web at scale. Vorhies notes that during this period, the legal frameworks governing data ownership were intentionally left ambiguous, allowing firms to build foundational models without compensating content creators.
- 2019 (The Whistleblower Intervention): Vorhies leaks 950 pages of Google internal documents, revealing that the company was using "human raters" to manipulate search outcomes and blacklist content deemed "fringe" or "fake news." This established a precedent for how tech platforms manage and curate the "truth."
- 2022 (The Lock-in): Following the public release of ChatGPT, AI companies began aggressively moving to limit web crawling for third parties. Companies like Reddit and X (formerly Twitter) locked down their APIs, effectively ensuring that only the largest AI cartels—who had already ingested the data—could maintain high-performing models.
- 2024–2025 (The Regulatory Capture): The U.S. government shifts from a passive observer to an active partner, with initiatives like the $500 billion "Stargate" project, which seeks to build massive, government-backed infrastructure to support the very companies that have already monopolized the data.
Altman’s "Utility" Vision and the Privatization of Intelligence
The strategic goal of this concentration is best summarized by OpenAI CEO Sam Altman. During a high-profile appearance at BlackRock in Washington, D.C., Altman articulated a future where intelligence is treated as a basic utility—analogous to electricity or water.
"People will buy it from us on a meter," Altman stated, a comment that has drawn sharp criticism from technology ethicists and privacy advocates alike. Vorhies contends that this business model is inherently predatory. By charging users for access to data that was originally collected from them—or from the public commons—these firms are engaging in a form of "data rent-seeking."
If intelligence becomes a metered utility, the power dynamic between the information producer (the human) and the information distributor (the AI company) is permanently inverted. The public is effectively paying to access its own collective knowledge, repackaged and filtered through the black-box algorithms of a handful of corporate entities.
The "Lawfare" Against the Digital Commons
A central pillar of Vorhies’ argument is the systematic destruction of free alternatives to the AI cartels. He points to the recent legal crackdown on digital repositories like Anna’s Archive and Z-Library as evidence of "lawfare"—the use of the judicial system to eliminate competition under the guise of intellectual property protection.
In late 2024, Anna’s Archive—which positions itself as the largest truly open library in human history—was ordered to pay $300 million in damages for scraping content. Similarly, the FBI’s 2022 seizure of Z-Library signaled a shift in federal enforcement priorities. Vorhies argues that these sites were the "Libraries of Alexandria" of our digital age. By destroying them, the state is not protecting creators; it is clearing the path for corporate monopolies to become the sole gatekeepers of knowledge.
This pattern aligns with observations made by economic analysts like David Dayen, author of Monopolized. Dayen has documented how dominant firms use regulatory hurdles and litigation to create "moats" around their businesses. When Senator Amy Klobuchar writes in her book Antitrust that innovation is stifled by a lack of competition, she describes the exact market conditions Vorhies claims are currently being engineered in the AI sector.
The Global Dimension: Competition from the East
While the U.S.-based cartels attempt to consolidate their hold on the market, they are increasingly facing pressure from international competitors. Vorhies highlights the rise of Chinese AI firms, specifically those behind the DeepSeek and Qwen models, as a potential disruptive force.
The emergence of the DeepSeek-R2 system, which utilizes architecture distinct from standard Western models, has forced a recalibration among global tech leaders. Vorhies posits that these "other cartels" from abroad may eventually force the U.S. firms to lower prices to stay competitive. In his view, a globalized, fragmented market of AI cartels is arguably safer for the public than a singular, state-backed domestic monopoly, as it prevents any one entity from holding absolute sway over the digital information flow.
Implications: The Need for Transparency
The most alarming aspect of the current trajectory is the opacity of the models themselves. Vorhies draws a parallel between his 2019 revelations regarding Google’s blacklisting practices and the current state of AI. He argues that we are currently operating in a "black box" environment where decisions regarding information curation, political bias, and content suppression are being made by algorithms that the public cannot audit.
"We are going to have to crack these models open and look inside," Vorhies asserts. His call for transparency is not merely about proprietary technology; it is about the civic right to understand how the tools that shape our perception of reality are functioning. If these models are "fair," they should be subject to independent, third-party audits—a prospect that current AI firms vehemently oppose.
The Stargate Initiative and the Future of Infrastructure
The U.S. government’s commitment of $500 billion to the "Stargate" initiative represents the culmination of this public-private partnership. By building national-scale AI data centers, the state is effectively subsidizing the infrastructure for the same firms that have been accused of monopolizing the data.
This raises profound questions about the future of the internet. Will the internet remain a decentralized network, or will it become a series of "walled gardens" owned by a few, government-blessed corporations? If the infrastructure is provided by the government but the content is owned by a cartel, the concept of a free and open internet may effectively cease to exist.
Conclusion: Reclaiming the Commons
The allegations made by Zach Vorhies paint a bleak picture of a future where human knowledge is commodified and guarded by a government-backed elite. Whether one views his perspective as a necessary warning or a critique of the inevitable realities of large-scale technological progress, the underlying issue remains: who owns the sum of human knowledge?
As the AI industry continues to grow, the tension between corporate monopolization and the public interest will only intensify. Without a concerted push for radical transparency and a re-evaluation of current antitrust and intellectual property laws, the "metered utility" model may become the permanent reality for global information access. The challenge for the coming decade will be to determine whether we can build AI that serves the public good without destroying the very foundations of the digital commons that made it possible.
