Discussions about the modern web often begin with concentration. A small number of technology companies account for a significant share of online search, cloud infrastructure, advertising, social media, mobile operating systems and AI services. From a user perspective, much of the Internet appears to pass through the platforms of a handful of firms.
That observation is broadly accurate. At the same time, it can obscure a more complicated structural reality.
Beneath the visible layer of consumer platforms, the web has continued to diversify across infrastructure, protocols, software ecosystems, identity systems and governance models. While some markets have become more concentrated, others have become increasingly distributed. The result is a digital environment that is less centralized than surface-level observations may suggest.
Understanding that distinction requires separating the applications people use from the systems that make those applications possible.
Layers Instead of a Single Platform
The web is not a single technology. It is a collection of interconnected layers with different economic incentives and governance structures.
At the application layer, network effects often encourage concentration. Social platforms become more valuable as more users join. Marketplaces benefit from larger buyer and seller communities. Search engines improve through extensive indexing and user interaction.
Infrastructure follows different dynamics.
The Internet continues to rely on open standards developed through organizations such as the Internet Engineering Task Force and the World Wide Web Consortium. These standards define protocols rather than products. Multiple organizations can implement them independently while remaining interoperable.
This distinction matters because protocol-level openness allows competition even when application markets become concentrated. A dominant browser, cloud provider or communication platform still depends on shared technical foundations that remain broadly accessible.
Cloud Computing Created Both Concentration and Distribution
Cloud infrastructure is often cited as evidence of centralization. A relatively small number of providers operate much of the world's hyperscale computing capacity.
Public reporting consistently identifies providers such as Amazon Web Services, Microsoft Azure, and Google Cloud as the largest participants in this market.
Yet cloud adoption has also reduced barriers for smaller organizations.
Before large-scale cloud platforms became widely available, deploying Internet services often required purchasing physical servers, maintaining data centers and investing heavily before reaching users. Today, startups, research organizations, governments and open source projects can launch globally without owning physical infrastructure.
This creates an apparent contradiction.
Infrastructure ownership has become concentrated, while service creation has become more decentralized. More organizations can build Internet services even though many rely on common infrastructure providers.
Those two developments occur simultaneously rather than excluding one another.
Open Source Has Shifted Where Control Exists
Software development increasingly depends on open source components.
Operating systems, programming languages, databases, container technologies, web frameworks and machine learning libraries frequently originate as collaborative projects involving companies, universities, nonprofits and independent maintainers.
Many commercial products build upon software that remains publicly available.
This does not eliminate concentration. Large companies often contribute significant engineering resources and influence the direction of important projects.
However, open source changes the nature of dependency.
Organizations adopting open software generally retain greater flexibility than users of entirely proprietary systems. Forking, independent governance and community maintenance create alternative paths when commercial priorities diverge from community interests.
The result is a more distributed innovation model than proprietary software ecosystems alone would produce.
AI May Increase Diversity as Well as Dependence
Artificial intelligence introduces another example of competing centralizing and decentralizing forces.
Training frontier foundation models requires substantial computing resources, specialized hardware, engineering expertise and capital investment. These constraints naturally favor large organizations.
Inference and deployment present a different picture.
Open weight models, smaller specialized models, local inference and hardware improvements have expanded the range of organizations capable of deploying AI without building frontier-scale training infrastructure.
Rather than producing a single dominant architecture, AI increasingly operates across multiple deployment models.
Some organizations prioritize proprietary cloud-hosted systems. Others combine commercial APIs with open models. Still others deploy specialized models entirely within private infrastructure for regulatory, security or operational reasons.
The technology therefore supports both concentration and diversification depending on which layer of the ecosystem is examined.
Regulation Encourages Multiple Paths
Public policy has also become an important factor in shaping decentralization.
Competition authorities in several jurisdictions have increased scrutiny of digital markets. Privacy regulation has altered data collection practices. New legislation addresses interoperability, platform responsibilities and digital competition.
These developments do not necessarily reduce the scale of large technology companies.
Instead, they often encourage greater interoperability, clearer governance or reduced switching costs between competing services.
According to public regulatory documentation in the European Union and other jurisdictions, many recent digital policy initiatives focus on preserving competitive conditions rather than replacing existing market leaders.
The long-term effects remain uncertain because regulatory outcomes depend on implementation, enforcement and technological change.
Geography Is Becoming More Diverse
For much of the commercial Internet's history, innovation appeared heavily concentrated within a relatively small number of regions.
That pattern has become less absolute.
Cloud regions now operate across dozens of countries. Internet exchange points continue expanding globally. Regional technology ecosystems increasingly develop software, AI models, cybersecurity products, payment systems, and digital services tailored to local legal and economic conditions.
Data sovereignty requirements have also encouraged greater geographic distribution.
Organizations handling sensitive information increasingly evaluate where data resides, how infrastructure is governed and which legal frameworks apply.
This does not fragment the Internet into isolated networks, but it does encourage a more geographically distributed digital infrastructure than existed during earlier stages of Internet growth.
Centralization Often Reflects Convenience
Many forms of concentration arise because centralized platforms reduce complexity.
Users typically prefer fewer accounts, integrated services, unified interfaces and familiar ecosystems. Developers often choose mature cloud platforms because they reduce operational overhead. Businesses frequently adopt widely supported software because integration costs are lower.
These incentives create concentration through coordination rather than exclusivity.
At the same time, improvements in open standards, containerization, interoperability, APIs and portable software reduce the technical cost of moving between providers compared with earlier generations of Internet services.
The practical outcome depends on switching costs rather than theoretical openness alone.
A system may remain technically open while becoming economically difficult to leave. Conversely, a concentrated market may still support meaningful competition if interoperability remains strong.
Looking Beyond the Surface
The future structure of the web is unlikely to be defined by a simple movement toward either complete centralization or complete decentralization.
Different layers respond to different incentives.
Consumer platforms benefit from network effects that encourage scale. Infrastructure providers benefit from economies of scale. Open standards encourage interoperability. Open source encourages distributed innovation. Regulation shapes competitive conditions. AI introduces both resource concentration and deployment diversity.
These forces interact continuously rather than moving in a single direction.
Viewed through the lens of consumer platforms alone, the web appears increasingly centralized.
Viewed across protocols, infrastructure, governance, software development and deployment models, the picture becomes more balanced.
The modern web remains a network of competing layers, overlapping institutions and evolving technical standards. Its future will likely be shaped less by whether centralization exists than by where it exists, how it is governed and how easily alternative paths remain available.
