AI is moving quickly from a development assistant to something far more ambitious: An operational actor. Agents can now generate code, propose changes, and even take action across infrastructure and applications. But while AI capabilities are accelerating, most platforms are still architected for humans to glue together fragmented systems.
This gap is where the hype begins to fall apart.
What’s missing isn’t more AI.
It’s platforms that can safely expose intent, context, policy, and operational state in a way both humans and machines can speak the same language. This is the problem the Intelligent Control Plane is meant to solve.
Rather than treating AI as an external tool bolted onto dashboards and runbooks, the Intelligent Control Plane evolves the control plane itself, unifying declarative state, actual state, policy, knowledge, and intelligence behind consistent APIs.
That sounds abstract, but parts of this model are already running in production today.
Control Planes as Contracts Between Teams
One of the foundations of an Intelligent Control Plane is deterministic control, leveraging Kubernetes’ API and reconciliation model to create predictable execution paths through clear APIs and policy enforcement.
At Allianz Technology, platform teams operate more than 1,000 Kubernetes control planes across a large organization. Their challenge wasn’t provisioning infrastructure; it was enabling development teams to consume platform capabilities safely, independently, and at scale.
The solution was to treat Kubernetes APIs as explicit contracts between teams.
Infrastructure teams expose capabilities through well-defined APIs. Development teams consume those APIs just like any other internal service, with clear ownership, versioning, documentation, and expectations. The platform enforces boundaries, while teams retain autonomy.
This approach eliminates ambiguity:
- What does the platform provide?
- What is stable versus evolving?
- Who owns failures and changes?
These API contracts form the deterministic layer of the control plane, the stable foundation required before any intelligence can be safely introduced.
From Alert Fatigue to Intelligent Control
Where deterministic control provides safety, intelligence begins to reduce toil.
At Millennium bcp, one of Portugal’s largest banks, platform teams faced escalating alert fatigue and long mean time to resolution across a regulated, multi-cloud environment. The goal wasn’t to hand control to an AI system, but to make operations more adaptive without sacrificing auditability.
The result was an AI-enhanced control plane built on Kubernetes and Crossplane.
Using LLM-powered composition functions, alerts are triaged automatically, and common remediation paths are executed within defined policy boundaries. Workload-aware algorithms assist with scaling decisions while every action remains observable, explainable, and compliant.
This is not autonomous infrastructure in the abstract.
It’s intelligent assistance layered onto a deterministic control plane, operating safely in production.
Why the Control Plane is the Right Place for AI
Kubernetes’ real innovation was never containers; it was the control loop: Desired state, actual state, and continuous reconciliation.
That same model provides the substrate for intelligent control.
AI does not replace reconciliation. It augments it. Intelligence can help determine what the desired state should be or how to respond when reality diverges, while the control plane ensures safe execution, policy enforcement, and recoverability.
This is why serious platform teams are embedding AI into control planes rather than bolting it onto external systems. Intelligence belongs where decisions already happen.
Less Hype, More Control
The Intelligent Control Plane is not about replacing engineers or skipping governance. It’s about evolving platforms so that both humans and agents can operate effectively, using the same APIs, policies, and sources of truth.
When platforms unify state, policy, and knowledge, AI becomes practical. Without that foundation, it remains hype.
Learn More at KubeCon + CloudNativeCon EU Amsterdam
These ideas aren’t theoretical. They’re being applied today by teams operating at real scale, under real constraints.
To see how Crossplane and the Intelligent Control Plane show up in production, join these sessions at KubeCon + CloudNativeCon EU 2026:
- API-Driven Infrastructure as Code: Kubernetes APIs as the Contract Between Teams
Florian Hopfensperger (Allianz Technology) and Yury Tsarev (Upbound) share how Allianz built API contracts between platform and application teams across 1,000+ Kubernetes control planes.
- From Alert Fatigue to Self-Healing: Building AI-Enabled Control Planes in Banking
Nuno Guedes (Millennium bcp) and Yury Tsarev (Upbound) walk through how AI-enhanced control planes reduce operational toil while remaining auditable and compliant in a regulated environment.
If you’re looking past AI hype and toward platforms that actually work in production, these real-world stories are a good place to start.