Mend AI

Mend AI is a security tool that finds and governs the AI components inside a company. It builds an AI bill of materials, called an AI-BOM, by scanning for AI models, frameworks, and datasets across an organization, including tools developers adopted without formal approval. It scores system prompts for injection risks, runs adversarial red-team tests against conversational AI to check for data leakage and bias, and applies runtime guardrails that check inputs and outputs against policy. It maps its checks to the OWASP Top 10 for large language model applications, the NIST AI Risk Management Framework, and ISO/IEC 42001. It is made by Mend, also known for open-source dependency scanning. Note that the vendor URL in some catalog records is incorrect; confirm details at Mend's own site.

This fits a company rolling out its own AI features, or one where employees adopt outside AI tools without formal review, that needs a way to find and govern that usage. A practical starting point is the AI-BOM discovery scan to see what AI components already exist before writing new policy. Mend has published a price near $300 per contributing developer per year for its AI product line, separate from its broader application security bundle near $1,000 per developer per year; confirm current packaging directly, since AI security offerings are still evolving quickly. The main risk is treating discovery as a one-time project instead of an ongoing scan, since new AI tools and models get adopted continuously.