Zscaler, OpenAI Unite to Secure Enterprise AI Deployments

Zscaler has joined OpenAI's Trusted Access for Cyber program, gaining access to security-tuned frontier models including GPT-5.4-Cyber. The partnership integrates these models into Zscaler's Zero Trust Exchange, enabling AI-powered vulnerability detection, secure software development, and red teaming capabilities. For enterprise customers, the collaboration addresses growing AI security risks including prompt injection, data exfiltration, and misconfigured model endpoints. Zscaler's AI Red Teaming platform uses OpenAI models to simulate multimodal attacks, generate automatic remediation steps, and analyze AI agents across organizations, helping businesses deploy AI at scale with greater confidence.