September 3rd, 2025 // Did someone forward you this newsletter? Sign up to receive your own copy here.

Who pays for the future of the web?

The business models of the web can be traced through a recurring cycle of unbundling and rebundling. Those were the terms that Netscape CEO Jim Barksdale used when he was taking Netscape public in the 1990s.

At the time, the World Wide Web unbundled the walled gardens of AOL, Prodigy, and CompuServe, replacing them with open-source software and protocols.

In the early 2000s, companies such as Facebook (now Meta), Google, and Amazon rebundled decentralized content into unified, centralized platforms.

Around the same time, CDs were unbundled into MP3s, then rebundled into streaming services like Napster and Spotify. Movies and TV shows were unbundled from traditional cable packages and reassembled into streaming platforms like Netflix.

“The digital revolution promised opportunity for all,” Frank McCourt, Founder of Project Liberty, said. “Instead, we’ve witnessed the unprecedented concentration of our data and power in the hands of a few.”

Today, the pendulum swings again. AI companies like OpenAI, MetaAI, Perplexity, and Anthropic are rebundling vast amounts of internet data into chatbot interfaces that scrape and consolidate information. The chatbot is becoming the primary interface to the modern web.

This shift will disrupt the entire web. With web traffic down, the ad-based business model that has funded much of the web for two decades is at risk. Media companies are filing lawsuits over unconsented scraping, while AI companies counter with fair use claims.

In this newsletter, we explore why the web needs a new business model, the dynamics driving this change, and what those models could be.

// Why AI is driving a new business model for the web

As journalist Casey Newton wrote in Platformer last month, “The entire structure of the web—from journalism to e-commerce and beyond—is built on the idea that webpages are being viewed by people.”

Yet, as we explored a few weeks ago with the rise of AI browsers, AI companies envision an entirely new approach to interacting with the web: Rather than individuals browsing site by site, AI mediates the experience through a single interface, centralizing access and influencing how information is discovered and delivered.

The fundamental unbundling-rebundling dynamic looks like this:

1. AI companies scrape the web—often without permission or compensation to content developers and owners.
2. They use that data to create new, self-contained experiences for users, reducing the need to visit the original sources.
3. Those original sources—from The New York Times to niche recipe blogs—lose traffic and, with it, ad revenue. The Economist reported global monthly search-engine traffic is down 15% year over year; TripAdvisor is down a third, WebMD by half, and Wikipedia by 8%.
4. With traffic and ad-based revenue declining, the economic incentives for producing high-quality content weaken.
5. AI’s dependence on scraped content creates a self-destructive cycle as the quality of the content it relies on becomes harder to sustain.

Without intervention, this becomes a tragedy of the commons for the internet. Barry Diller, chairman of IAC, warned: “If all the world’s information is sucked up in this maw, and then essentially repackaged in declarative sentences…there will be no publishing; it is not possible.”

But there is room for optimism. New business models are emerging that could sustain content creation while supporting innovation.

// The emerging new business models

We’re at the start of the next rebundling. New business models, like those outlined in Project Liberty Institute’s report on the Fair Data Economy, are emerging that balance data rights with innovation and growth. 

Consider the following models:

License & syndication. Media companies license content directly to AI firms. In 2024, News Corp signed a $250 million deal with OpenAI to use its content for training and queries. The New York Times struck a deal with Amazon while continuing its lawsuits against Microsoft and OpenAI. The Associated Press, Financial Times, and Dotdash Meredith have inked similar agreements.

Pay-per-crawl & API access. Cloudflare’s pilot program lets publishers decide whether to allow, block, or charge AI crawlers each time they request content (via the public protocol HTTP response code 200).

Attribution-based revenue sharing. Perplexity shares ad revenue with publishers when its chatbot uses their content, with partners like the Los Angeles Times and Adweek. Zendy compensates academic publishers based on citation frequency in AI responses.

Digital asset business models. On the individual level, people are beginning to sell their personal data, and a range of digital ownership models are emerging, including the Frequency blockchain.

Closed content ecosystems. Paywalls and subscriptions protect content from scraping while generating direct revenue. The New York Times has doubled digital subscribers since 2020 to nearly 12 million. Platforms like Substack offer smaller creators similar protection.

Community-supported content. Wikipedia’s and Signal’s donation model and Patreon’s creator memberships show the enduring power of direct audience support. Patreon has 290,000 creators, who collectively earn $25 million every month from their fans.

No single model will dominate; hybrid approaches are likely. Yet without careful design, a more privatized, fragmented internet could emerge—where valuable content retreats behind walls. Business model innovation alone isn’t enough. Regulation is needed to enforce intellectual property rights and ensure AI companies operate within established law.

// The next chapter of the web

In the last 30 years, the internet has reinvented itself multiple times—a cycle of unbundling and rebundling again and again. We’re at the beginning of the next cycle, but this time the stakes are higher: AI’s rebundling threatens the original promise of the internet as a decentralized place to create and share information.

Paul Fehlinger, Director of Policy, Governance Innovation & Impact at Project Liberty Institute, who has led the Fair Data Economy Task Force and spearheads PLI’s engagement with LPs and VCs on the future of the AI economy, said:

“AI is the biggest economic transformation of our time. Fixing the broken business models of the web is not just a policy challenge—it’s one of the greatest entrepreneurial opportunities of this decade.”

Fehlinger continued, “Investors who back better business models that can scale in the AI race won’t just generate strong returns—they’ll help build the next generation infrastructure of a better internet and more resilient societies.”

The next cycle of unbundling and rebundling awaits.

Project Liberty news & updates

// UNGA side-event on Scalable Alternative Governance and Business Models for a Fair Data Economy
September 16th, 4pm - 6pm ET, New York City
Project Liberty Institute is hosting a side-event at the United Nations General Assembly (UNGA) in New York City with speakers from UN agencies and leading think-tanks. The gathering will examine innovative models and case studies, including data cooperatives, privacy-centered governance models, and data infrastructure solutions aligned to the SDGs. Express your interest here to participate in the event.

📰 Other notable headlines

// 🤖 More people are turning to general-purpose chatbots for emotional support, according to an article in The New York Times. At first, Adam Raine, 16, used ChatGPT for schoolwork, but then he started discussing plans to end his life. (Paywall).

// 🔞 Selfies, government IDs and AI are all being used by companies in efforts to adhere to new age-verification laws and regulations aimed at protecting children, according to an article in The Wall Street Journal. (Paywall).

// 📣 An article in Noema Magazine argued that defending democracy in the digital age will require moving beyond the focus of fighting online misinformation. (Free).

// 🏛 Mississippi’s new age assurance law is putting decentralized social networks to the test, according to an article in TechCrunch. (Paywall).

// 👤 Some industry leaders and observers have a new idea for limiting mental health tragedies stemming from AI chatbot use: They want AI makers to stop personifying their products, according to an article in Axios. (Free).

// 🧠 While large language models like ChatGPT can impress with their speed and fluency, relying on them too heavily can make us slower, duller, and less able to think for ourselves. An article in Project Syndicate examines how to use AI without losing our minds. (Paywall).

// 🤔 We’re used to algorithms guiding our choices. But an article in The New Yorker asks, when machines can effortlessly generate the content we consume, what’s left for the human imagination? (Paywall).

// 🏆 Time Magazine introduced the TIME100 AI 2025, a list of the 100 most influential people in AI. Big shout out to Project Liberty collaborators Latanya Sweeney, Refik Anadol, and Alex Bores. (Free).

Partner news

// A guide for starting a tech-focused nonprofit 

Jim Fruchterman’s new book, Technology for Good: How Nonprofit Leaders Are Using Software and Data to Solve Our Most Pressing Social Problems, is now available from MIT Press. Drawing on stories from more than sixty nonprofits, the book explores how to start a nonprofit tech company focused on maximum social good, not profits. 

// Debate: Should the U.S. be ruled by a CEO dictator?

September 4th at 7pm ET

What if the U.S. abandoned democratic governance for a CEO-style dictator — someone running the country like a high-performing company? This idea is gaining momentum in certain policy circles and is also endorsed by prominent Silicon Valley figures. Curtis Yarvin, an anti-democracy theorist and tech entrepreneur, will argue yes. E. Glen Weyl, Co-Founder, RadicalxChange Foundation, will argue no. Watch here.

// AI’s impact on the future of work

September 29th at 3pm ET

A new landmark study from Stanford, Canaries in the Coal Mine?, is sparking conversation about how AI is reshaping the U.S. labor market. Join the authors on September 29 for a free virtual seminar exploring the paper’s findings and what it means for the future of work. Register here.

What did you think of today's newsletter?

// Project Liberty builds solutions that help people take back control of their lives in the digital age by reclaiming a voice, choice, and stake in a better internet.

Thank you for reading.