Cloud + Infrastructure

What enterprise autonomy can do now, where the economics work, why programs stall, and the architecture and operating model required to scale AI agents safely.

NetApp expanded AI Data Engine metadata discovery across its own and third-party storage and deepened its Commvault recovery integration, signaling a larger shift from copy-heavy AI pipelines toward governed data activation and resilience where enterprise data already resides.

Anthropic’s Claude 5.5 is now available through AWS at essentially unchanged token prices. The bigger shift is that AI economics are moving toward cost per accepted task—while the control plane around the model may prove more durable than the model itself.

NVIDIA’s Open Agent Safety Platform exposes the deeper problem with autonomous AI: useful agents are becoming privileged software, and prompts are not a security boundary.