Open source software has undergone a profound transformation in the last few years. Once viewed as a fringe alternative driven by developer idealism, it has quietly become the backbone of modern AI infrastructure. The shift is not about diminishing importance, but rather about a new kind of centrality – one where control is exercised through contribution and standardization rather than ownership. As the AI industry races to deploy powerful models, the tools and platforms that make those deployments possible are overwhelmingly open source. From Kubernetes orchestrating machine learning pipelines to OpenTelemetry providing observability, open source is no longer a nice-to-have; it is the substrate on which the AI economy is built.
Control through code
While headlines focus on the latest closed-source AI models, the underlying infrastructure is being shaped by collaborative development. The Cloud Native Computing Foundation (CNCF) now hosts more than 230 projects with over 300,000 contributors worldwide. Its 2025 survey found that 98% of organizations have adopted cloud-native techniques, and 82% of container users run Kubernetes in production. GitHub's Octoverse report for 2025 reveals 1.12 billion contributions, more than 180 million developers, and a record 518.7 million merged pull requests. The Apache Software Foundation remains robust with 9,905 committers working across 295 projects and issuing 1,310 software releases in fiscal year 2025. These numbers are not just growth; they represent a consolidation of effort around the layers that matter most for AI.
The companies driving this contribution surge are not acting out of altruism. Red Hat led all CNCF contribution activity in 2025 with 194,699 contributions, followed by Microsoft with 107,645, and Google with 91,158. Independent contributors remained significant, landing fourth at 52,404. This pattern has remained consistent over the past decade, indicating a long-term strategic investment. The message is clear: open source is where vendors set defaults, normalize interfaces, and shape the operational assumptions everyone else must adopt. It is less about openness for its own sake and more about control over the layers where ecosystems harden into standards.
Who gives, and why?
Red Hat’s dominance in CNCF contributions is directly tied to its product strategy. OpenShift, Red Hat’s Kubernetes-centric platform, depends on the health of the Kubernetes ecosystem. Contributing heavily ensures that the platform remains central to enterprise deployments. Similarly, Microsoft’s position as second-largest contributor reflects a dramatic shift from its historical hostility toward open source. Microsoft’s investments in OpenTelemetry, which saw a 39% rise in commits in 2025 and a contributor base growth from 1,301 to 1,756, are strategic moves to set observability standards. Splunk and other top contributors also participate to shape a market they depend on. This is not charity; it’s a land grab around critical infrastructure.
Cilium, the eBPF-based networking project, illustrates how boring infrastructure becomes exciting when it sits at the intersection of networking, observability, and security. After joining CNCF, Cilium saw contributing companies rise 90% (from 533 to 1,011) and individual contributors jump from 1,269 to 4,464. Google, Datadog, and Cloudflare all expanded their contributions as the project matured. Cilium is precisely the kind of infrastructure that determines whether AI workloads are governable, visible, and efficient. As AI models become distributed and latency-sensitive, such projects become mission-critical.
Nvidia provides another telling example. Despite having enormous financial resources, the company ranked 14th in Kubernetes contributions in the past two years with 5,892 contributions. It also open-sourced the KAI Scheduler, a Kubernetes-native GPU scheduler that came out of its Run:ai acquisition. Nvidia has described itself as a key contributor to Kubeflow. Instead of simply buying influence, Nvidia is investing in the scheduling, orchestration, and workflow layers that determine how effectively its chips are used in real-world AI systems. This approach leverages developer communities rather than cash payouts.
An essential supporting actor
The trajectory is clear: AI is making open infrastructure more important than ever. According to the CNCF, 66% of organizations hosting generative AI models now use Kubernetes for some or all inference workloads. The foundation explicitly calls Kubernetes the de facto operating system for AI. While this may be self-serving, it reflects reality. Kubernetes and Kubeflow are increasingly central to training and inference systems. Organizations do not want to build their future on opaque, inescapable infrastructure they cannot inspect or influence. Open source provides that transparency and control.
Historical context reinforces this shift. The early open source movement was driven by a desire for freedom and community governance. But as corporate involvement grew, the focus moved from ideology to utility. Today, open source is the mechanism through which competing companies jointly standardize the layers beneath their products. This is not a betrayal of the original vision; it is an evolution. The same dynamics that made Linux the dominant server operating system are now playing out in AI infrastructure. Companies contribute to shape the defaults, and those defaults become the foundation for entire industries.
Consider the example of OpenTelemetry. This project has become the de facto standard for observability data collection. Microsoft and Splunk invest heavily because they sell products that consume and analyze telemetry data. By ensuring OpenTelemetry is widely adopted, they ensure their own products remain relevant. Similarly, Google’s contributions to Kubernetes and Kubeflow align with its cloud business. Google Cloud runs on Kubernetes, and Kubeflow provides a path for AI workloads to run on that platform. Every contribution is a strategic bet on the future architecture of computing.
For individual developers, the implications are significant. The days of contributing to open source solely for peer recognition or moral satisfaction are fading. Instead, contributions are increasingly tied to career advancement and industry influence. Developers who master projects like Cilium, OpenTelemetry, or Kubeflow are positioning themselves at the center of AI infrastructure. The skills required to contribute to these projects are in high demand. As the infrastructure becomes more complex, the barrier to entry rises, but the rewards for those who navigate it grow accordingly.
The transformation also affects smaller organizations and startups. They can no longer afford to ignore the open source ecosystem. Building proprietary alternatives to Kubernetes or OpenTelemetry is impractical because the network effects are too strong. Instead, startups must build on top of these platforms, contributing where possible to gain visibility and influence. This creates a virtuous cycle: more contributions improve the platforms, which attract more users, which generates more contributions. The ecosystem becomes self-sustaining, but it also becomes more corporate-driven.
Critics argue that this corporate dominance undermines the original spirit of open source. They point to incidents where companies have re-licensed projects or used open source as a marketing tool. While there is some truth to these concerns, the overall trajectory is positive. Open source projects today are better funded, more secure, and more widely used than ever before. The corporate influence ensures long-term maintenance and innovation. Moreover, the governance models of foundations like CNCF and ASF provide checks and balances that prevent any single company from taking complete control.
Looking ahead, the intersection of AI and open source will only deepen. New projects are emerging to address the unique challenges of AI: model versioning, data lineage, experiment tracking, and ethical compliance. Many of these projects are open source from day one. The trend toward open infrastructure for AI is driven by the same forces that drove the adoption of Linux and Kubernetes: the desire for portability, transparency, and community-driven innovation. As AI becomes embedded in every industry, the open source projects that support it will become as critical as the models themselves.
In this new landscape, the role of the open source contributor has evolved. Developers are no longer just coding in their spare time; they are making strategic decisions that shape the future of technology. Companies that understand this invest in developer relations and contribution programs not as a charitable expense but as a strategic imperative. The result is a more professional, more capable, and more influential open source ecosystem. The romance of the early days may be gone, but in its place is something arguably more important: a foundation for the next generation of computing.
Source: InfoWorld News