Google announced a comprehensive suite of AI-powered security enhancements at the Google Cloud Security Summit 2025, introducing advanced capabilities designed to protect AI ecosystems while leveraging artificial intelligence to strengthen organizational defenses.
The announcements span three major areas: securing agentic AI deployments, empowering security operations centers with autonomous agents, and expanding cloud security controls across Google’s unified platform.
Advanced AI Agent Protection Capabilities
Google’s Security Command Center now features expanded AI agent inventory and risk identification capabilities, enabling automated discovery of AI agents and Model Context Protocol (MCP) servers.
This preview functionality helps security teams identify vulnerabilities, misconfigurations, and high-risk interactions across their AI agent ecosystem.
The company extended Model Armor’s in-line protection to Google Agentspace prompts and responses, providing real-time safeguards against runtime threats, including prompt injection, jailbreaking, and sensitive data leakage.
These protections work alongside specialized posture controls tailored for AI agents in both Agentspace and Agent Builder, ensuring agents adhere to security policies and organizational standards.
Building on Mandiant’s threat intelligence, Google introduced new threat detections specifically for AI agents, helping organizations identify anomalous and suspicious behavior on agent assets.
The Sensitive Data Protection service has been expanded to monitor Vertex AI and AI-related assets in BigQuery and CloudSQL, now capable of inspecting images for sensitive data like barcodes and license plate numbers.
Autonomous Security Operations and Alert Management
Google’s vision of an agentic Security Operations Center (SOC) materialized with the preview release of the Alert Investigation agent.
This autonomous system performs dynamic investigations following Mandiant analysts’ best practices, enriching events, analyzing command-line interfaces, and building process trees to reduce manual effort while accelerating response times.
The agent provides verdict-based recommendations for human intervention, representing a significant step toward automated alert triage and investigation workflows.
Google Security Operations also introduced new dashboards with native SOAR data integration, offering enhanced visualization and analysis capabilities for security data.
Expanded Cloud Security Controls
Google’s Trusted Cloud platform received substantial updates, including Compliance Manager in preview, which unifies policy definition, control configuration, and evidence generation for streamlined security and compliance workflows.
The Data Security Posture Management service now integrates natively with BigQuery Security Center, allowing data professionals to monitor security posture without context switching.
Identity and Access Management improvements include an IAM role picker powered by Gemini AI for least-privilege access recommendations and re-authentication requirements for sensitive actions like billing account updates.
These enhancements demonstrate Google’s commitment to making security an enabler of business innovation while maintaining comprehensive protection across AI deployments.
Find this Story Interesting! Follow us on Google News , LinkedIn and X to Get More Instant Updates
