Google Cloud Unveils AI-Powered Security Ally to Relieve Overworked Teams and Protect Businesses’ AI Initiatives
Google Cloud envisions a future where Artificial Intelligence (AI) serves as an indispensable aid to overburdened security teams, rather than an additional burden. This vision was unveiled at the Security Summit 2025.
The focus is on utilizing AI to safeguard organizations while concurrently fortifying their own AI initiatives against cyber threats. Given the escalating reliance on AI agents in businesses, these agents now represent a significant frontier for security concerns.
To establish AI as a reliable defender, its own environment must first be secure. Google Cloud is enhancing its AI Protection solution within the Security Command Center to achieve this. New functionalities, set to debut in preview soon, will automatically identify all AI agents and servers within your environment. This provides security teams with a comprehensive view of their entire AI agent ecosystem, facilitating the detection of vulnerabilities, misconfigurations, and risky interactions.
Real-time protection is also being bolstered. Model Armor’s in-line protection is being extended to cover prompts and responses within Agentspace, thereby preventing threats such as prompt injection and data leaks as they occur.
To ensure AI agents comply with company security policies, new posture controls are being implemented. Additionally, new threat detections powered by intelligence from Mandiant and Google Cloud will enable security teams to more effectively identify and respond to unusual or suspicious behavior from their AI assets.
Google’s most visionary announcement pertains to the concept of an “agentic security operations center (SOC)”. This system envisions AI agents collaborating to manage threats, automate alert investigations, and even assist engineers in creating new detections to bridge security gaps.
The initial step towards this goal is the introduction of the Alert Investigation agent, which is currently in preview. This tool operates like a junior analyst, independently examining security events, analyzing command-line activity, and constructing process trees based on the proven methods of Mandiant’s frontline experts. The agent offers its assessment on alerts and suggests subsequent actions for human analysts, promising to reduce manual work and accelerate response times.
In Google Security Operations, the new SecOps Labs provides early access to potent capabilities, many of which are powered by Gemini AI. New dashboards that consolidate security orchestration, automation, and response (SOAR) data are now available, offering a more coherent perspective on an organization’s security posture.
The platform’s core, the Trusted Cloud, is also undergoing enhancements:
By integrating AI into the very essence of its offerings, Google Cloud aims to create a foundation where security propels business objectives and equips defenders with the tools necessary to confront the challenges of this new era.
