Which is a limitation of cloud hosting for AI?

• A. Less control, potential vendor lock-in, performance concerns, and security/privacy risks
• B. Full control and no risks
• C. Guaranteed best performance
• D. No security concerns

Answer: (A)

The main idea is recognizing the practical limitations and trade-offs of using cloud hosting for AI workloads. Cloud environments offer scalability and managed services, but they come with real constraints. You have less direct control over hardware, software stacks, and rapid incident response, since these are managed by the provider. This can limit customization and create dependency on vendor practices for patching and security.

Vendor lock-in is another real concern: relying on provider-specific APIs, tools, and data formats can make moving workloads or data to another environment costly and complex. Performance can also be less predictable in the cloud due to multi-tenant resources, potential latency, and regional differences, which matters for AI workloads that are sensitive to timing and throughput. Security and privacy risks are part of cloud usage too, since data and models live outside your own premises, requiring careful configuration and governance to meet regulatory and internal standards.

The other options imply absolute control with no risks, guaranteed top performance, or no security concerns, which don’t reflect how cloud hosting actually behaves in practice.