Enterprise infrastructure teams have spent the last decade moving workloads into Kubernetes. If it runs in a container — apps, APIs, batch jobs, or data pipelines — it belongs in the cluster. Kubernetes is now the default way to run production workloads.
But desktop and app delivery have stayed outside this model. Legacy virtual desktop infrastructure (VDI) was built for another time. It used pre-allocated VM pools and custom management tools that don’t fit modern container platforms. This split means teams have to handle different tools, scaling methods, and runbooks.
Platform engineers who know Kubernetes well must switch gears when dealing with desktop infrastructure problems. Yet, the desire for Kubernetes-native workspace delivery is growing as organizations mature their container platforms. They want to manage everything—apps and desktops—using the same tools and workflows.
Containerized workspaces offer better session isolation than VM-based desktops. Sessions are ephemeral, isolated within containers, and end without leaving persistent state behind. This level of isolation is tough to match with traditional VDI.
Kasm Workspaces Brings Desktops into Kubernetes
Kasm Workspaces is built to use Kubernetes as the control plane for workspace orchestration and delivery. It comes with production-grade Helm charts and tested upgrade paths. Its backend architecture has been validated across many deployments.
Kasm includes an RDP Gateway to let users reach Windows and Linux VMs through Kubernetes. It supports horizontal session scaling, declarative configuration via Helm, and namespace-level isolation. It also exports metrics and supports rolling builds for smooth updates.
Organizations can deploy Kasm in the same Kubernetes cluster as their app platform. This means workspace sessions are managed with the same operational tools and pipelines. Kasm sessions are ephemeral, with controlled network egress, and can be managed through GitOps pipelines.
Kasm is useful for contractor or third-party access. It can scale sessions during engagement periods and reduce persistent access after. When combined with NVIDIA MiG Multi-Instance GPU support, Kasm can allocate fractional GPU resources to isolated workspace sessions.
This Kubernetes-native approach lets platform teams manage workspace infrastructure using the same dashboards and pipelines as their applications. It cuts overhead, improves consistency, and eliminates the mental context switches between desktop and app infrastructure.
Organizations still running legacy VDI alongside modern infrastructure should consider moving to Kubernetes-native solutions. This transition simplifies operations and aligns desktop delivery with container platform standards.
AI Governance Demands Zero Trust and Faster Security
Alongside Kubernetes adoption, AI governance and security have become top priorities. AI systems, especially agentic AI, are growing more than fivefold. These systems act faster than human attackers, pushing security teams to respond quicker.
Andre Durand, CEO of Ping Identity, says, “The rise in desire to use agents right now, and the speed of agentic, is highlighting the need to move faster on the principles of zero trust.” He emphasizes that each AI agent should have its own identity and should never impersonate a human.
Durand explains that security is no longer about whether someone is logged in. “It’s your next action that we care about,” he says. Permissions must be checked at every action, not just at login.
Shared secrets like API keys embedded in source code are risky. They should be replaced with secure authentication methods. Enforcement of zero trust policies can happen at API gateways and agent gateways that inspect requests before granting access.
AI systems acting without documented human approval pose audit risks. Every AI decision should be logged with the decision point, who reviewed it, and the outcome. Connecting risk, policy, vendor, and training data helps enterprises manage AI risks comprehensively.
Preparing for AI Regulation and Compliance
Regulators are stepping up enforcement too. The European Union’s AI Act threatens fines up to €35 million or 7% of global annual turnover for deploying prohibited AI systems. The UK’s FCA also provides AI governance guidance through DORA.
These enforcement timelines are not vague. They are specific and approaching fast. Organizations that build AI compliance architectures now will be ready for these enforcement windows.
One key takeaway is that the future of AI compliance isn’t about who uses AI first. It’s about who can account for every decision AI informs. This means logging, auditing, and continuous verification must become standard practice.
The rise of Kubernetes-native workspaces and stricter AI governance shows one thing clearly. Infrastructure and security are merging into more unified, automated, and accountable platforms. This shift will shape how enterprises run apps, workspaces, and AI in the years ahead.
Based on
- From intent to enforcement: Lessons from operating Kubernetes controllers at scale — thenewstack.io
- The desktop infrastructure problem that Kubernetes finally solves | VentureBeat — venturebeat.com
- Zero trust must now move at agent speed | VentureBeat — venturebeat.com
- Companies Are Building AI Workforces, but Many Still Need to Deploy Governance Systems | AP News — apnews.com
- The future of AI-powered compliance is defensible | The Independent — independent.co.uk
