Tego AI Finds Claude Tag Slack Integration Can Trigger Unauthorized Enterprise Actions

Tego AI disclosed research showing Anthropic’s Claude Tag Slack integration can be triggered by literal “@Claude” text in non-intentional messages (e.g., bots, webhooks), potentially allowing the agent to perform unauthorized enterprise actions such as data retrieval and resource deletion; Tego recommends least-privilege connections, avoiding untrusted channels, retaining logs, and independent runtime authorization, while Anthropic has classified the submission as informative and disputed that default configurations are vulnerable.