Enterprises are rethinking cloud strategies because AI workloads are changing the requirements for performance, cost, and control.

According to the Information Services Group (ISG), more than 80% of enterprises are revisiting their cloud plans specifically to better support AI. The main drivers are:

• Operational resilience and consistency: AI services need to be reliable and performant across different environments.
• Cost pressure: GPU-intensive AI workloads can quickly increase cloud spend, so organizations are looking for more cost-conscious models.
• Vendor lock-in concerns: As AI usage grows, leaders want flexibility to avoid being tied too tightly to a single cloud provider.
• Sovereignty and governance: Enterprises want more control over where data lives and how AI is governed, especially for long-term AI plans.

In practice, this means companies are redesigning their operating models, investing in GPU-enabled architectures, and adopting hybrid AI operating models that can span multiple environments.

Modern hybrid cloud for AI is less about a single location for workloads and more about maintaining control across a mix of environments.

ISG notes that popular hybrid approaches now blend:

• Public cloud for elastic, on-demand AI compute
• Private cloud for sensitive workloads and tighter governance
• Colocation for custom or high-density GPU infrastructure
• Edge environments for low-latency AI close to users or devices
• Sovereign cloud for stricter data residency and regulatory needs

As ISG’s cloud delivery lead for the Americas puts it, hybrid cloud has become less about where workloads run and more about how enterprises maintain control across these diverse environments.

To support this, organizations are:

• Investing in GPU-enabled architectures and distributed data platforms
• Adopting hybrid AI operating models that span on-premises and multiple clouds
• Looking for unified platforms that combine observability, automation, and financial management in one framework

The goal is to reimagine cloud as a coordinated operating fabric for AI, rather than a single destination.

Cost, transparency, and sovereignty are now central to how enterprises design AI-ready cloud strategies.

Cost and transparency:

• AI workloads and GPU-heavy infrastructure are driving up spend across private cloud, Kubernetes, edge, and sovereign environments.
• Organizations are demanding better cost transparency and optimization tools so they can see and manage AI-related costs end to end.
• ISG highlights that enterprises increasingly expect providers to unify operations, cyber recovery, and cost management rather than treat them as separate disciplines.

Sovereignty and control:

• Sovereignty is a major factor behind diversifying cloud offerings, as enterprises want more control over infrastructure, data residency, and AI governance.
• This is especially important for long-term AI plans in regulated or data-sensitive industries.

Market response:

• Vendors are building AI-optimized infrastructure and services to meet these needs.
• “Neocloud” providers focused on AI-ready compute are projected by Gartner to capture 20% of a $267 billion AI cloud market by 2030.
• Recent moves, such as Apple opening access to private cloud compute and a compute-as-a-service offering from Blackstone and Google, show how providers are reshaping offerings around AI demand.

For enterprises, the practical takeaway is that AI strategy and cloud strategy are now tightly linked, with cost discipline, transparency, and sovereignty built in from the start.