Zscaler Inc. has announced a strategic partnership with OpenAI, aiming to transform its zero-trust security platform into an AI accelerator. This collaboration will enable Zscaler to strengthen its own defenses and assist customers in deploying artificial intelligence safely at scale.
Zscaler is now part of OpenAI’s Trusted Access for Cyber (TAC) program, granting access to specialized AI models like GPT-5.4-Cyber. These tools will be integrated into Zscaler's Zero Trust Exchange and its AI Red Teaming and SecOps offerings.
Internally, Zscaler is embedding GPT-5.4-Cyber and Codex-style models into its security architecture and software development lifecycle. This provides developers with AI-driven security reviews, identifying and remediating vulnerabilities earlier in the process, a concept known as "shift left" security.
Externally, these AI capabilities will power Zscaler's AI Red Teaming services and Managed Detection and Response (MDR) investigations. The AI Red Teaming platform uses OpenAI's multimodal models to simulate sophisticated attacks against customer AI applications, identifying weaknesses like prompt injection and tool abuse.
Furthermore, the partnership aims to accelerate customers' AI initiatives by addressing new attack surfaces introduced by enterprise AI deployments, such as LLM platforms, agent frameworks, and sensitive data integration. Zscaler's zero-trust architecture ensures these AI systems are only accessible to authenticated users and workloads.
This alliance allows enterprises to experiment with AI more rapidly and establish consistent AI security baselines across the organization. By integrating AI-driven findings into development and response workflows, Zscaler aims to help customers deploy AI faster and more securely, boosting confidence in their AI roadmaps.