GOOGLE has launched unique AI agent identities as part of its Gemini Enterprise Agent Platform, a hub to manage agentic AI workflows for both Google-made and external agents. Each agent can be assigned a unique cryptographic ID that is mapped back to defined authorization policies that are traceable and auditable, according to Google Cloud’s CEO Thomas Kurian, speaking at Cloud Next 26.
The platform also includes Agent Registry, which indexes every internal agent, tool and skill, and Agent Gateway, a dashboard to manage agent-to-agent and agent-to-tool connections with security policies and Model Armor protections to guard against prompt injection and data leakage, according to Google Cloud. Google Cloud said the new features bring zero trust verification to every agent and at every orchestration step.
At the same event, Google Cloud unveiled Agent Anomaly Detection and the Agent Security dashboard, built to identify unusual behaviour in real time and map relationships between AI agents and models, with threat detection for hostile activity and vulnerabilities scanning within Google Cloud Platform environments, according to Google Cloud.
The updates were announced alongside three new security-focused agents for cyber professionals, including Threat Hunting and Detection Engineering agents, both available in preview, with a Third-Party Context agent coming soon to preview. Finally, Google claimed its Triage and Investigation agent, introduced in April 2025, processed over five million alerts in the last year, reducing a typical 30-minute manual analysis to 60 seconds, and a new dark web intelligence feature is now available in preview.