SentinelOne announced on November 2 an integration with Anthropic’s Claude Cowork that puts its security tooling around AI agents running on employee devices. The company’s press release describes the offering.
A SentinelOne agent runs inside the virtual machine that Claude Cowork uses, which gives security teams visibility into what the AI agent does there. The company’s Behavioral AI engine is applied to that activity to flag anomalous behavior and what it calls agent drift. Coverage spans macOS and Windows devices.
IT teams switch the integration on through device management, so organizations decide when and where it applies. The release is written in future tense and gives no general availability date, so rollout timing remains unclear.
The announcement reflects a wider push among endpoint vendors to treat autonomous AI agents as a new class of workload that needs the same monitoring as laptops and servers.