Google Cloud unleashes AI security agents as $32B Wiz deal powers agentic defense strategy

Google Cloud Shifts to AI-Led Defense Strategy

At Google Cloud Next 2026 in Las Vegas, Google unveiled a comprehensive agentic defense portfolio that marks a fundamental shift in how organizations protect their digital infrastructure1

1

. Francis deSouza, COO and president of security products at Google Cloud, articulated the transformation clearly: "It is very clear that we have moved from a human-led defense strategy, to a human-in-the-loop defense strategy, to an AI-led defense strategy that's overseen by humans"2

2

. This evolution reflects a stark reality—human analysts simply cannot process threats fast enough to counter AI-powered adversaries who now compress attack windows from hours to as little as 22 seconds, according to the Mandiant M-Trends 2026 report4

4

.

Three New AI Security Agents Join the Fight

Google Cloud introduced three new AI security agents within its Google Security Operations offering, designed to enable machine-speed cyber defense5

5

. The Threat Hunting agent proactively searches for novel attack patterns and stealthy behaviors that bypass traditional defenses, using intelligence from Google Threat Intelligence and Mandiant best practices3

3

. The Detection Engineering agent identifies security coverage gaps in IT environments and continuously creates new detection rules based on findings2

2

. A Third-Party Context agent, coming soon to preview, will enrich existing security workflows using third-party data5

5

. All three agents are currently in preview mode, building on security-specific agents announced last year.

Proven Results from Existing Agentic Systems

The Triage and Investigation agent, announced at last year's conference, is now generally available and has already demonstrated substantial impact2

2

. Over the past year, this agent processed more than 5 million alerts, reducing typical 30-minute manual analysis times down to approximately 60 seconds by leveraging Gemini3

3

. Google claims its dark web-crawling threat intelligence agents can "analyze millions of daily external events with 98% accuracy to help elevate only the threats that truly matter to your organization"1

1

. This performance matters because the expanding cyber attack surface and AI-powered adversaries can deploy attacks at speeds "well beyond the powers and abilities of mortal humans," forcing defenders to adopt AI armies of their own1

1

.

The $32 Billion Wiz Acquisition Powers Integration

Google's parent company Alphabet completed its acquisition of Wiz last month for $32 billion in an all-cash transaction—the largest cybersecurity acquisition ever and the biggest purchase in Alphabet's history1

1

. This figure exceeds Canada's entire military defense budget and approaches Israel's military spending, signaling what the tech giant views as nation-state-level urgency1

1

. Wiz, founded in 2020, built its reputation on an ability to find vulnerabilities in networks and software platforms. Under Google Cloud, the Wiz platform has expanded support to include Databricks, AWS Agentcore, Gemini Enterprise Agent Platform, Microsoft Azure Copilot Studio, and Salesforce Agentforce5

5

.

Securing AI-Native Development and Shadow AI

Wiz announced new capabilities specifically for securing AI-native development, including an AI Bill of Materials (AI-BOM) to help mitigate the risk of Shadow AI2

2

. As developers create AI applications using various skills, SDK libraries, models, and MCP servers, Wiz provides security teams with a complete inventory of components used to build these applications, according to Yinon Costica, Wiz co-founder and vice president of product2

2

. The platform now integrates with Lovable, running security scanning to surface vulnerabilities, secrets, and misconfigurations directly to developers during coding—not after deployment2

2

. This addresses a critical concern raised by Morgan Adamski, U.S. cyber, data and tech risk leader at PricewaterhouseCoopers: employees building and deploying agents outside formal governance structures4

4

.

Building Custom Agents and Future Implications

Google Cloud customers can now build their own security agents with remote Google Cloud model context protocol (MCP) server support for Google Security Operations, now generally available2

2

. Users can access the MCP server client directly from the Google Security Operations chat interface, though this feature remains in preview2

2

. Google also launched its Gemini Enterprise Agent Platform for building, orchestrating, and governing AI agents, with functionality including agent identity, agent gateway, and Model Armor runtime protection5

5

. DeSouza emphasized that Google's "full AI stack"—developing chips, models, and every layer in between—differentiates it from competitors, allowing the company to "take advantage of them on day one and use those most sophisticated models available to create the agentic fleet"2

2

. Charles Carmakal, CTO of Mandiant, noted that while "with AI, you can scale tremendously, well beyond what humans are able to do," there will always be risks associated with this approach4

4

. As organizations race to deploy agentic fleets that operate at "infinite scale," the question remains whether human oversight can effectively govern AI systems making split-second security decisions3

3

.