Cloud Security Alliance Report Highlights Risks in Managing AI Identities as Non-Human Entities

A report by the Cloud Security Alliance titled The State of Non-Human Identity and AI Security reveals that most organizations manage AI-related identities as non-human identities (service accounts, API keys, chatbots). This approach perpetuates existing weaknesses in Identity and Access Management (IAM) systems, including credential sprawl, unclear ownership, and inconsistent lifecycle controls. The document emphasizes that the massive integration of AI identities exacerbates these risks without specific adjustments to security policies.