Claude Code 2.1.310, released October 23, brings the Monitor tool to AWS Bedrock, Google Vertex AI, and Azure AI Foundry. It also adds a new Safety telemetry event that logs every blocked tool call, plus /import commands to migrate settings from Gemini CLI, Cursor, or Codex in one step.
Anthropic opened the Claude Security public beta to Enterprise customers on October 21, 2026. Rather than scanning once and moving on, it monitors repositories continuously, re-analyzes changes as they happen, and tracks each vulnerability until it is fixed. Team-informed threat models and API/MCP integrations are included.
Anthropic launched two enterprise API betas on October 20β21, 2026. The Claude Tag API lets admins control @Claude behavior per Slack channel through code. The User Profiles API ties each API request to a specific end user, so safety enforcement can target individuals instead of the entire organization.
Anthropic launched Claude Duet as a research preview. It sends a daily morning brief like a messenger, monitors your connected apps in the background even when your laptop is closed, and keeps working on tasks β a standby AI agent built for proactive work.
Anthropic released Claude Haiku 5.5 on October 7. The context window grew to 1M tokens, max output doubled to 128K, and it's the first Haiku-class model with adjustable effort β letting you tune how deeply the model thinks per request.