Cloud Compliance: Bridging Hyperscalers and Neoclouds
The rapid escalation of AI workloads is fundamentally rewriting the economics and architecture of enterprise infrastructure. While global hyperscalers have long dominated the conversation, a new breed of "neocloud" providers is rapidly gaining traction.
Offering the high-density compute, transparent pricing and data sovereignty demanded by GPU-heavy and heavily regulated applications, neoclouds are successfully breaking the hyperscaler-first mindset.
However, this transition is not without friction. Operating a sprawling, multi-cloud estate without a unified strategy often leads to a fragmented security posture, shadow IT policies and entirely unpredictable costs.
The challenge for today's enterprise is no longer whether to adopt multiple cloud providers, but how to orchestrate them without descending into operational chaos.
To unpack this critical industry shift, Data Centre Magazine sits down with Dmitry Panenkov, CEO and Founder of cloud management platform emma.
Dmitry explores why bridging the gap between hyperscalers and neoclouds requires a radical rethinking of infrastructure governance. He outlines the hidden dangers of decentralised adoption and explains why a centralised, automated control plane is the ultimate key to delivering AI at scale – transforming a complex multi-cloud web into a secure, cost-efficient and cohesive fabric.
Why is the rise of neoclouds reshaping the economics and architecture of AI infrastructure? What does this mean for enterprises?
Neocloud providers are reshaping AI infrastructure because they offer specialised capabilities that address the cost, performance and regulatory pressures created by modern AI workloads. While hyperscalers excel at global scale, elasticity and managed services, they are not always the most efficient option for GPU-intensive training, sovereign deployment and predictable pricing.
Neoclouds fill this gap by providing high‑density compute and more transparent economics, particularly for GPU-heavy workloads and regulated industries. As a result, enterprises are moving away from a “hyperscaler only” mindset. Instead, they’re adopting a strategic mix of hyperscaler breadth and neocloud specialisation. The shift is no longer about whether to use neoclouds, but how to operate across multiple providers without introducing operational complexity.
How does decentralised cloud adoption without unified controls lead to shadow policies, inconsistent security posture and unpredictable workload behaviour?
When organisations adopt multiple clouds without unified governance, each team tends to configure environments independently, resulting in shadow policies and inconsistent operating practices.
Neoclouds fill this gap by providing high‑density compute and more transparent economics, particularly for GPU-heavy workloads and regulated industries. As a result, enterprises are moving away from a “hyperscaler only” mindset. Instead, they’re adopting a strategic mix of hyperscaler breadth and neocloud specialisation. The shift is no longer about whether to use neoclouds, but how to operate across multiple providers without introducing operational complexity.
How does decentralised cloud adoption without unified controls lead to shadow policies, inconsistent security posture and unpredictable workload behaviour?
When organisations adopt multiple clouds without unified governance, each team tends to configure environments independently, resulting in shadow policies and inconsistent operating practices.
Identity, networking, encryption and tagging standards drift because no central function ensures they remain aligned across providers.
This creates a fragmented security posture in which workloads inherit different defaults and protections depending on where they run.
Over time, the organisations lose predictability. Workloads scale differently, logging behaves inconsistently and infrastructure and cost patterns vary widely across clouds.
Without a unified governance model, the multi-cloud strategy becomes difficult to secure and prone to operational surprises.
Why is centralised governance – spanning costs, performance, security and compliance – the key to enabling a “multi-cloud for AI” strategy that is flexible without becoming chaotic?
Centralised control is essential for multi-cloud AI because it provides a single operational layer that ensures consistency across all environments, regardless of provider.
By defining policies for cost, performance, security and compliance in one place, organisations can ensure that every workload is deployed and operated according to the same standards. This allows teams to choose hyperscalers or neoclouds based on business intent, such as locality or cost efficiency, without creating cloud-specific exceptions or silos.
Unified governance also automates policy enforcement, ensuring that guardrails are applied continuously rather than retroactively. This not only reduces risk but also streamlines operations by abstracting away provider differences.
With this model in place, multi-cloud becomes a deliberate, flexible strategy for delivering AI at scale rather than a chaotic collection of independently managed environments.
Looking to the future, where do you see the data centre sector heading?
The future of cloud is one where hyperscalers, neoclouds, sovereign providers, on-premise systems and edge environments operate as a cohesive, interconnected fabric.
Workloads will move fluidly between environments based on business, regulatory and performance needs, not on vendor lock-in or tooling limitations.
Networking will be inherently multi-cloud, interoperability will improve and governance will rely increasingly on automation. Centralised control planes will play a critical role by ensuring that distributed architectures remain secure, cost-efficient, compliant and predictable.
In this future, organisations gain genuine freedom of choice without sacrificing control, and multi-cloud becomes the default design pattern for AI-driven enterprises.