A recent breach at Mercor has been linked to a supply-chain attack on LiteLLM, a dependency used by the company, allowing attackers to gain access to internal environments and harvest credentials on a large scale1. This incident highlights the growing concern of AI system exposure and limited visibility, making it challenging for organizations to detect and respond to such attacks. The breach reportedly exposed sensitive data and source code, emphasizing the need for organizations to reevaluate their AI dependencies and implement robust security measures. The attack on LiteLLM is a significant concern, as it can be used as a stepping stone for further attacks on other organizations that rely on the same dependency. This incident matters to practitioners because it underscores the importance of securing AI dependencies and monitoring for potential supply-chain attacks to prevent similar breaches and protect sensitive information.