Wiz Partners with Google Cloud to Build Machine-Speed Defences for AI Era

The security problem has not changed. The scale of it has. And Wiz is taking notice.

Misconfigurations, over-privileged identities, unpatched vulnerabilities—these have been the core challenges for security teams for decades. What AI has done is give those problems a megaphone. Autonomous agents now pull levers in sensitive environments at a speed that human defenders simply cannot match. A minor oversight that once stayed localised can, in a hyper-connected AI ecosystem, become a systemic vulnerability within minutes.

Wiz’s response to this is the AI Application Protection Platform—AI-APP—which the company introduced last month at RSAC and expanded significantly at Google Cloud Next. The core idea is to move security from human speed to machine speed, covering the entire AI development and deployment lifecycle rather than addressing individual gaps after the fact.

What Is Actually New

The announcements span three areas that matter to anyone managing cloud and AI infrastructure.

The first is platform coverage. Wiz has added support for Databricks and expanded visibility across AI studio platforms including AWS Agentcore, Google’s Gemini Enterprise Agent Platform, Microsoft Azure Copilot Studio, and Salesforce Agentforce. The practical value here is understanding how AI agents interact with data and connected tools—and therefore where new exposure paths are being created, often without security teams knowing they exist.

The second is a new Red Agent, now in public preview. It functions as an AI-powered simulated attacker—continuously probing for complex, logic-driven vulnerabilities in real time rather than waiting for scheduled assessments. In an environment where advanced AI models are making zero-day exploitation easier, having an intelligent system that reasons about your attack surface before adversaries do is the kind of proactive capability that traditional security tooling was not built to provide.

The third is the Technology Intel Centre, which aggregates cloud and AI provider updates—new feature releases, migration notices, end-of-life announcements—and maps them directly onto an organisation’s existing infrastructure. The result is not just awareness that something has changed in the cloud; it is clarity about exactly how that change affects your risk posture and, for organisations using Wiz Cloud Cost, your spending.

The AI Coding Problem

Perhaps the most immediately relevant development for security professionals is what Wiz is doing around AI-assisted coding. Wiz Research analysed real applications built with AI coding tools and found that 20% contained material security issues—broken access controls, unprotected data endpoints. The tools that accelerate development are also accelerating risk.

Three new capabilities address this directly. The AI Bill of Materials automatically inventories AI frameworks, models, and IDE extensions—including GitHub Copilot, Gemini Code Assist, and Cursor—mapping them onto the Wiz Security Graph. Post-code-generation hooks inject security guardrails before AI writes the code and scan the output before it reaches source control. And new pre-built remediation skills feed AI agents with code-to-cloud context from the Security Graph, enabling issues to be identified and fixed directly within developer workflows.

Wiz Spans Edge to Core

Wiz has also extended its Security Graph to cover the cloud edge through integrations with Cloudflare, Akamai, Vercel, and Google Cloud Apigee—bringing DNS configurations, edge policies, frontend deployments, and API ecosystems into a unified risk view alongside core cloud infrastructure.

The point is straightforward: a risk at the edge is still a risk. If it is not visible in the same context as the workloads it connects to, it will not be prioritised appropriately.

Security at AI scale requires machine-speed response. Wiz is building the infrastructure to make that possible.