AI Agents and Non-Human Identities Creating Critical Security Gaps, Report

Keeper Security's RSA Conference report warns that many organizations lack visibility and proper controls for non-human identities (NHIs) such as service accounts, API keys, and AI agents: 46% grant AI tools access to sensitive systems, 76% lack consistent privileged-access rules, only 28% can see all NHIs, and over 40% reported security incidents involving machine credentials in the past year, indicating a widespread risk of credential-based breaches.