The Agentic Web and Original Sin
Conceptual representation of the evolving web landscape
In his recent Stratechery article, Ben Thompson presents a compelling analysis of how AI agents are fundamentally transforming the structure of the web and its underlying business models. Thompson draws fascinating parallels between today's emerging "agentic web" and the early evolution of the commercial internet, suggesting we may be witnessing a similar paradigm shift in how users interact with online content and services.
The Rise of the Agentic Web
Thompson defines the "agentic web" as an emerging internet layer where AI agents act as intermediaries between users and content, fundamentally changing how information is discovered, filtered, and presented. Unlike traditional search engines that provide links for human evaluation, these agents actively interpret, synthesize, and act upon information on behalf of users.
This shift represents a profound change in the web's architecture. As Thompson notes:
"The original sin of the web was not advertising, but rather the fundamental disconnect between how humans process information and how the web delivers it. AI agents bridge this gap by serving as cognitive extensions, processing vast amounts of information in ways that align with human needs rather than technical constraints."
Business Model Disruption
Perhaps the most significant insight from Thompson's analysis is how the agentic web threatens established business models. When AI agents become the primary interface between users and content, traditional advertising-based models face existential challenges. Agents that filter out ads or bypass ad-supported content entirely could undermine the economic foundation of many web properties.
Thompson argues that this disruption parallels earlier web transitions, particularly the shift from desktop to mobile. Just as mobile apps disintermediated websites, AI agents may disintermediate both websites and apps, creating a new layer of abstraction between users and content providers.
The New Gatekeepers
A critical concern in Thompson's analysis is the emergence of new gatekeepers. As AI agents become primary interfaces for web interaction, the companies controlling these agents gain unprecedented power over information discovery and access. This raises important questions about market concentration, algorithmic transparency, and the potential for new forms of rent-seeking behavior.
Thompson suggests that regulators and industry stakeholders should be proactive in addressing these concerns:
"The window for establishing governance frameworks for the agentic web is rapidly closing. Without thoughtful intervention, we risk recreating the same power imbalances that characterized earlier internet eras, but with even greater concentration and less transparency."
Adaptation Strategies
For content creators and businesses, Thompson outlines several potential adaptation strategies for the agentic web era:
- Agent-optimized content: Creating content specifically designed to be processed and recommended by AI agents
- Direct agent relationships: Establishing formal relationships with leading agent platforms
- Subscription models: Shifting away from advertising toward direct user revenue
- Proprietary agents: Developing specialized agents for specific content domains or user needs
Critical Assessment
While Thompson's analysis is insightful, it may overestimate the speed of transition to an agentic web. Current AI agents still struggle with hallucinations, outdated information, and limited understanding of context. These limitations suggest that human-mediated information discovery will remain valuable for the foreseeable future.
Additionally, Thompson's analysis could benefit from deeper consideration of how different cultural contexts might shape agent adoption and use. The agentic web may evolve differently across global markets with varying regulatory approaches, digital literacy levels, and cultural attitudes toward AI.
Conclusion
Thompson's analysis of the agentic web provides a valuable framework for understanding one of the most significant transitions in internet history. By drawing parallels with earlier web evolutions, he helps contextualize current developments and anticipate potential challenges.
For businesses, content creators, and policymakers, the emergence of the agentic web demands strategic reconsideration of established practices and business models. Those who adapt effectively to this new paradigm may find significant opportunities, while those who resist change risk disintermediation and irrelevance.
As we navigate this transition, Thompson's call for thoughtful governance and business model innovation serves as an important reminder that technological change alone does not determine outcomes. How we collectively shape the agentic web will ultimately determine whether it enhances or diminishes human agency in the digital realm.
References
- Thompson, B. (2025). "The Agentic Web and Original Sin." Stratechery. https://stratechery.com/2025/the-agentic-web-and-original-sin/
- Altman, S. (2024). "Agents: A New Computer Interface." OpenAI Blog.
- Andreessen, M. (2023). "Why AI Will Save the World." a16z Future.
- Tufekci, Z. (2024). "The Algorithmic Public Sphere." The Atlantic.
- Zuboff, S. (2019). "The Age of Surveillance Capitalism." Profile Books.