Mend.io Introduces System Prompt Hardening for AI Security

Mend.io has launched System Prompt Hardening within its Mend AI platform to detect, score, and automatically remediate weaknesses in AI system prompts before applications enter production. The tool addresses security risks posed by hidden instructions in system prompts, which traditional application security (AppSec) tools do not fully mitigate. Powered by Mend.io’s proprietary AI Weakness Enumeration (AIWE) framework, it provides visibility into prompt logic, identifies vulnerabilities, and strengthens prompts to reduce exposure. The solution targets a growing concern in AI security by hardening prompts prior to deployment. No specific dates, CVE IDs, or numerical impact metrics were disclosed in the announcement.