Google Debuts AI & Cloud Security Upgrades at Summit
▼ Summary
– Google announced updates at its Cloud Security Summit 2025 to enhance AI security and enterprise defenses, including protections for AI agents and cloud security tools.
– New AI security features for Google Agentspace and Agent Builder include expanded agent inventory, risk identification, and proactive threat detection with a focus on balancing flexibility and control.
– Google introduced the Agentic SOC vision, featuring an alert investigation agent that automates triage and investigation to free human analysts for more complex tasks.
– Cloud security updates include new features in Security Command Center for compliance and threat management, plus identity and access management tools powered by Gemini.
– Data and network security enhancements cover expanded monitoring for AI assets, image inspection for sensitive data, and stronger protections in Chrome Enterprise for mobile devices.
Google has unveiled a comprehensive suite of security enhancements at its Cloud Security Summit, introducing powerful new tools designed to protect AI-driven workflows and reinforce enterprise cloud infrastructure. These updates focus on securing generative AI agents, streamlining security operations, and extending data protection across hybrid environments.
A major emphasis was placed on securing AI development, with significant upgrades to Google Agentspace and Agent Builder. These platforms now feature expanded agent inventory tracking, advanced in-line protection mechanisms, and more robust posture controls. According to a Google Cloud product lead, the goal is to embed security deeply into the development process without hindering innovation. The approach includes layered protections, detective policies for visibility and preventative controls to block high-risk actions by default. For specialized needs, core security features remain accessible via REST API, allowing developers to tailor defenses to specific use cases.
The concept of an Agentic SOC received considerable attention, outlining a future where AI agents work together to automate alert triage, investigation, and response. A new alert investigation agent, currently in preview, enriches security events and analyzes command-line activity using techniques adapted from Mandiant’s expert analysts. This system is designed to free human analysts from routine tasks, enabling them to focus on deeper investigative work and root cause analysis.
Beyond AI-specific tools, Google rolled out several cloud security improvements. Security Command Center now includes Compliance Manager, enhanced Data Security Posture Management, and detailed Risk Reports, all aimed at simplifying regulatory adherence and improving threat visibility. Identity and access management saw the introduction of a Gemini-powered IAM role picker, agentic IAM for automated identity provisioning, and mandatory re-authentication for sensitive operations.
Data protection capabilities were also expanded, with better monitoring of AI-related assets in BigQuery and CloudSQL, sensitive data detection in images, and the general availability of Autokey in Cloud Key Management. Network security upgrades include broader Cloud NGFW tagging and preview support for zero trust architectures in high-performance computing setups.
For enterprise browsing, Chrome Enterprise introduced stronger mobile protections for iOS and Android, incorporating URL filtering and new reporting tools to monitor security events across mobile devices.
(Source: HelpNet Security)
