Overview
- Anthropic has operationalized a Claude Compliance API that streams conversation content and activity event logs to third‑party security products so teams can monitor chats, uploads, user logins, admin actions, and configuration changes programmatically.
- The company announced integrations with 28 vendors including CrowdStrike, Palo Alto Networks, Microsoft Purview, Okta, Zscaler, Netskope, Cloudflare, Datadog, Fortinet, and Wiz so Claude data flows into existing DLP, SIEM, identity, e‑discovery, and observability systems.
- At its Code w/ Claude event in London Anthropic launched a public‑beta self‑hosted sandbox that runs Claude Managed Agents on customer‑controlled infrastructure so files, repos, and tool execution do not leave a customer perimeter.
- Anthropic also released a security‑guidance plugin for Claude Code that flags about 25 risky code patterns in real time and which the company says cut security comments on internal pull requests by roughly 30–40 percent during rollout.
- Those product moves build on Anthropic’s vulnerability work — its Mythos model flagged thousands of potential issues across open‑source projects and Claude Code Security claims 500+ previously unknown high‑severity findings — and aim to lower CISO friction for large Claude deployments.