ClawVault blocks AI data leaks

- ClawVault launched an AI security vault designed to block data leaks and prompt injection, acting as a proxy for OpenAI and Anthropic model access. - It offers fine‑grained controls to prevent PII exposure, enforce usage policies and manage AI access centrally for applications and agents. - The product frames itself as a governance layer for AI access, useful where teams need controlled, auditable model interactions. (x.com)