Cloudflare, Inc., a leader in cloud connectivity, has announced a collaboration with Wiz, a prominent company in cloud security and artificial intelligence—currently integrated into Google Cloud—with the goal of providing security teams with a unified solution to analyze and protect all AI applications within their environment. Through the integration of Cloudflare AI Security for Apps into Wiz Security Graph, organizations gain a much more comprehensive view of AI presence in their systems, along with effective tools for its protection.
The rapid deployment of AI features in companies is surpassing security teams’ control capacity. Every AI-based chatbot, co-pilot, or endpoint represents a potential attack vector, exposed to risks such as malicious instruction injection, sensitive data leaks, or misuse. The main challenge lies in identifying which solutions are active, ensuring they have appropriate security measures, and verifying their effectiveness. To achieve this, security leaders (CISOs) need a centralized source that helps them better understand and protect both their AI infrastructure and their cloud environment.
This integration enables organizations to eliminate blind spots related to AI across their infrastructure, promoting a safer adoption of these technologies. On one side, Cloudflare AI Security for Apps provides perimeter protection, securing AI endpoints against threats like instruction injection or unsafe content. On the other, Wiz’s AI Application Protection Platform (AI-APP) identifies and maps all AI applications, detecting potential vulnerabilities. By combining Cloudflare’s security rules with Wiz Security Graph, teams can prioritize risks based on their exploitability level. As a result, CISOs gain greater visibility into large language models (LLMs) within their web environments and access real-time controls to enforce security policies. Additionally, both Cloudflare and Wiz are provider- and model-independent, ensuring endpoint protection regardless of the LLM or cloud infrastructure used.
“AI is the most transformative technology we’ve seen in a generation, enabling countless functions. However, for most companies, it remains a mystery. When speaking with CISOs today, they point out struggles in balancing innovation with AI and tackling uncontrolled Shadow AI across their organization, as their legacy security tools are effectively useless at this level,” says Tom Evans, Cloudflare Partner Director. “The partnership between Cloudflare and Wiz helps solve this problem. Now, we offer a solution that allows rapid AI innovation without risking the exposure of the most sensitive data.”
“Security alignment isn’t just about reducing risk; it must also facilitate AI application development,” states Oron Noah, Vice President of Product, Extensibility, and Partnerships at Wiz. “By combining Wiz’s comprehensive visibility with Cloudflare’s perimeter protections, we address one of the main gaps in AI risk management. This partnership provides organizations with a unified view of their AI endpoints and shared risk context, helping them block threats like instruction injection and Shadow AI before they manifest.”
Cloudflare and Wiz’s partnership enables clients to:
- Identify Shadow AI: Detect all LLM endpoints across your web properties (including those deployed without security team involvement) and understand which are protected and which are exposed.
- Inspect AI traffic in real time: Cloudflare AI Security for Apps conducts detections on all requests to LLM endpoints to identify and mitigate PII leaks, instruction injection, and customized topics. These detections run in parallel across Cloudflare’s global network, adding no latency to AI traffic.
- Create a map of confidential data flows: Wiz maps data flows between AI applications, models, and data stores within Security Graph, providing security teams visibility into where sensitive data interacts with AI workloads to prioritize remediation efforts.
- Verify security measures: Wiz ensures AI deployments are protected by Cloudflare’s AI Security for Apps. If security measures are lacking or misconfigured, Wiz alerts teams to fix issues directly within the Cloudflare platform.
- Prioritize remediation: The integration highlights unprotected AI endpoints with access to sensitive data or production systems, allowing security teams to address the highest risks first.
This seamless integration requires no custom workflows or additional agents for deployment. Cloudflare’s detections run inline on their extensive and highly interconnected global network, enabling organizations to benefit from AI security without architecture changes or performance trade-offs.