Claude Managed Agents Memory public beta under managed-agents-2026-04-01 header

AI Impact Summary

Anthropic has released a memory feature for Claude Managed Agents in public beta, activated by the managed-agents-2026-04-01 header. This enables persistence of agent state across turns, allowing more coherent multi-turn dialogues but introduces data retention and privacy considerations that must be addressed in design, governance, and storage planning. Teams should review the 'Using agent memory' integration guide to understand how to enable, configure retention, and scope memory per agent, and prepare migration/testing plans for existing bots to switch to memory-enabled flows. Expect changes to latency and storage costs, and ensure that memory usage complies with data-handling policies and regulatory requirements.