HAIEC
Scans AI application code, runs adversarial tests against live endpoints, and generates signed compliance evidence
HAIEC is an AI security validation platform that combines three checks into one run: static analysis of application source code, authorized adversarial testing against live AI endpoints, and evaluation of technical controls against regulatory frameworks.
The static scanner covers prompt injection, missing authentication, tool abuse, RAG poisoning and tenant isolation, and exports SARIF. Runtime testing probes deployed endpoints for jailbreaks, system prompt extraction and data exfiltration. Results are packaged as SHA-256 signed artifacts mapped to SOC 2, ISO 42001, the EU AI Act, NIST AI RMF, OWASP LLM, NYC LL144, GDPR and HIPAA.
It integrates through a GitHub App and CI/CD hooks, with npm and PyPI packages. Aimed at teams that have to show an auditor or enterprise buyer how an AI system was tested, rather than assert that it was.
Pricing: Monthly subscriptions
HAIEC Alternatives
Explore 45 products in the Observability & Analytics category. View all HAIEC alternatives.
RAGAS
Open-source evaluation and testing framework for LLM and RAG applications
Guardrails AI
Open-source framework for adding input and output validators around LLM calls
NeMo Guardrails
NVIDIA toolkit for adding programmable guardrails to LLM conversational apps
Presidio
Microsoft open-source SDK for detecting and anonymizing PII in text and images
Work on HAIEC? Feature it at the top of Observability & Analytics.
Is your product missing?