Cloud computing has been sold as elastic, on-demand access to virtually unlimited resources. However, the rapid growth of data-intensive and artificial intelligence–driven workloads has strained the availability of certain types of computing, particularly specialized processors and region-specific capacity. As a result, customers (and their lawyers) are questioning whether compute resources will be available when needed.
What Is the Customer Actually Buying?
Cloud contracts often use the language of “capacity” without clearly defining it. From the customer’s perspective, the distinction between on-demand services, reserved instances, committed use discounts, and pre-purchased capacity blocks is critical. Each model carries different assumptions about availability and provider discretion.
Ambiguity around infrastructure requirements, geographic scope, or whether resources are shared or dedicated can undermine a customer’s expectations, even where the provider believes it has met its contractual obligations. Clear definitions are proving important in determining whether the customer is buying access to the resources its workloads actually require.
Capacity Assurance and Allocation Risk
Perhaps the most significant risk for customers lies in the gap between capacity “commitments” and capacity guarantees. Many cloud agreements stop short of firm assurances, instead obligating providers to use “commercially reasonable efforts” to make capacity available. At the same time, providers often retain broad rights to allocate resources across customers and to prioritize certain workloads. In periods of scarcity, these provisions can leave customers without meaningful recourse.
From a customer standpoint, it is helpful to consider whether capacity is guaranteed, the circumstances under which it may be reallocated, and the consequences if promised resources are unavailable.
Term Commitments and Flexibility Cost
Capacity commitments are inherently predictive, requiring customers to forecast future needs in a rapidly evolving technological environment. Long-term commitments can secure access to scarce resources, but they also increase the risk of over-buying, under-utilization, or technological obsolescence. Many cloud agreements impose financial penalties if usage falls short of committed levels, with limited rights to adjust or terminate.
When negotiating, customers should weigh the financial factors against contractual flexibility to enable the ability to scale up and down, shift regions or instance types, and allow for exit.
Pricing Considerations
Pricing provisions may obscure the true cost of capacity commitments. Prepayments, minimum spend obligations, and blended discounts can mask exposure to usage shortfalls or provider-initiated changes. Customers will want to carefully assess repricing rights, pass-through charges, and unilateral modifications to service descriptions that may erode the value of a capacity commitment over time. Capacity terms and pricing terms should be considered together to confirm that the economic bargain holds under both business-as-usual and stressed conditions.
Provider Change Rights and Evolving Infrastructure
Cloud providers may reserve broad rights to modify their services, including retiring instance types, upgrading infrastructure, or migrating workloads. While such changes may be operationally justified, they can directly affect capacity commitments. It is helpful for the customer to read the change rights carefully and, at a minimum, consider notice or transition protections, particularly where workloads depend on specific hardware or configurations.
Regulatory and Compliance Constraints
For regulated customers, capacity commitments often intersect with data residency, security, and audit requirements. Restrictions on where data may be processed can limit a provider’s ability to substitute capacity in other regions, while subcontracting arrangements may raise compliance concerns. Careful review of the regulatory constraints is critical to facilitating a customer’s ability to deploy workloads lawfully and consistently.
Force Majeure and Foreseeable Scarcity
Supply-chain disruptions and infrastructure shortages are increasingly foreseeable risks in the cloud market. Customers should scrutinize force majeure provisions to ensure that predictable capacity constraints are not broadly excused.
Conclusion
As cloud computing capacity becomes more constrained, cloud agreements may be designed to preserve provider flexibility while limiting customer certainty. Reviewing what capacity is promised, how it is protected, and what remedies apply if it is not delivered are critical contractual negotiation points.
How We Can Help
Morgan Lewis’s technology transactions, outsourcing, and commercial contracts lawyers regularly advise clients on complex technology deals, global sourcing strategies, and evolving regulatory risk. If you have questions about the topics discussed above or would like to learn more, please reach out to any member of our team.