Zscaler SPLX Dynamic Remediation is an AI security tool that addresses vulnerabilities in generative AI systems through automated prompt hardening and targeted remediation strategies. The platform identifies security risks in AI applications and generates hardened system prompts designed to reduce attack surfaces and prevent common vulnerabilities. The tool provides system prompt hardening capabilities that analyze existing prompts and generate more secure versions based on identified risks. Users can compare old and new prompts to see security improvements before deployment. The platform claims to reduce attack surfaces by 90% for specified GenAI risks and improve prevention of system prompt leakage by 97%. Dynamic remediation steps guide users through targeted fixes for specific risks uncovered during AI red teaming assessments. The platform tracks applied security measures and allows ongoing monitoring of security posture improvements. Users can reassess their AI systems after applying hardened prompts to validate security enhancements. The solution integrates with AI systems through REST API connections, conversational platforms, and various large language models. It requires no coding for setup and provides connectors for different AI deployment types. The platform is part of Zscaler's broader AI security offering following the acquisition of SPLX.