The centre of gravity has moved below the model
The most revealing change in open source is not that it has become less visible, but that it has become more foundational. At a moment when public attention is fixed on proprietary AI models and the companies behind them, the real strategic shift is happening deeper in the stack. Open source has not lost relevance in the AI era; it has moved into the layers where control, interoperability, and production readiness are decided. Kubernetes, observability frameworks, networking projects, and platform tooling may lack the glamour of frontier models, but they are increasingly the systems that determine whether AI can operate at scale.
Table of Contents
This helps explain why open source can appear quieter while becoming more consequential. The evidence cited from CNCF, GitHub, and the Apache Software Foundation points not to decline, but to consolidation around operational infrastructure. Cloud-native techniques are now nearly universal, Kubernetes is firmly established in production environments, and contribution activity across the broader software ecosystem remains immense. What changed is not the volume of open source, but the kind of power it now exerts. It is less a symbolic alternative to proprietary software than the substrate on which commercial AI and cloud services increasingly depend.
Corporate participation is no longer a side story
The contribution data also makes one point unmistakable: open source is being shaped heavily by major technology companies. Red Hat, Microsoft, and Google sit among the top contributors to CNCF projects, while independent developers remain important but no longer define the centre of gravity on their own. This does not mean open source has been hollowed out. It means the incentives have become clearer. Companies are investing engineering resources upstream because the projects they influence are the same ones their products rely on downstream.
That matters because it changes the meaning of contribution itself. Too often, open source work is still described as if it were mainly an act of goodwill or community obligation. The reality presented here is far more strategic. Companies contribute because open source is where interfaces harden into defaults, where operational assumptions become standard practice, and where long-term leverage is won. Influence over these layers is not proprietary in the traditional sense, but it is still power: the power to shape the environment that everyone else must build within.
AI is making infrastructure projects more strategic, not less
The article’s examples make that logic concrete. Red Hat’s continuing strength in Kubernetes reflects the fact that OpenShift depends on the Kubernetes ecosystem remaining central. Microsoft’s rising influence around OpenTelemetry reveals a similar pattern in observability, where defining how systems are measured and understood can shape an entire market. Cilium’s growth shows how networking, security, and observability have converged into infrastructure that becomes indispensable once workloads are distributed, latency-sensitive, and costly to run.
Nvidia’s behaviour is especially telling because it highlights where AI is pushing open source next. Rather than relying solely on its dominance in hardware, the company is contributing to the orchestration and workflow layers that determine how efficiently GPUs are used in practice. Its work around Kubernetes-related tooling, scheduling, and Kubeflow suggests that the future value of AI will not be captured only by those who build the chips or the models, but also by those who help define the systems that allocate, govern, and optimise them. That is a critical distinction in an era when inference and training costs are becoming central to AI strategy.
Open source is becoming the control plane for production AI
This is why Kubernetes now appears so prominently in discussions of AI infrastructure. As organisations host and operate generative AI systems, they increasingly need tooling that can manage distributed workloads, enforce visibility, and support efficient scheduling across expensive computing resources. In that environment, open infrastructure becomes attractive not because it is ideologically pure, but because it is inspectable, adaptable, and difficult for any single vendor to monopolise entirely. AI is increasing the value of open source precisely because it raises the cost of depending on systems that users cannot meaningfully influence.
The broader conclusion is that open source has matured into something less romantic but more essential. It is no longer persuasive to describe it as a fringe movement or a moral counterweight to commercial software. The more accurate picture is that open source has become the place where cloud-native architecture is standardised, where observability is normalised, and where AI infrastructure is quietly being assembled. The spotlight may remain on models, but the durable power is accumulating in the open systems that make those models usable in the real world.
Author:
Jan Bielik
CEO & Founder of Webiano Digital & Marketing Agency
Source: How AI is changing open source
