IBM and CoreWeave co-design controls for agent workloads

# IBM and CoreWeave Team Up to Manage AI Agents Running in the Real World As AI systems move beyond just being trained to actually performing tasks—like writing code or using tools—companies need better ways to keep these systems organized and secure in their data centers. IBM and CoreWeave are working together to build control systems that prevent AI agents from interfering with each other or accessing data they shouldn't, similar to how IT departments isolate different company departments' computer systems. This is becoming essential because running multiple AI agents simultaneously is becoming common, but the infrastructure to safely manage them hasn't caught up.