Entersoft AI Application Security Testing (AIAST) is a security testing framework designed to identify vulnerabilities in AI-powered applications, specifically those built on Large Language Models (LLMs) and Retrieval Augmented Generation (RAG) systems. The framework extends traditional application security testing methodologies to address AI-specific attack surfaces. AIAST encompasses two specialized testing verticals: RAG AST for securing retrieval augmented generation pipelines, and LLM AST for applications powered by large language models. The framework tests for vulnerabilities including prompt injection, data poisoning, hallucinations, sensitive data leakage, agent abuse, and supply chain risks in AI systems. The testing methodology includes architecture review and threat modeling, attack surface discovery, adversarial testing, vulnerability validation, remediation and retesting, and governance mapping. Testing covers AI data flows, trust boundaries, third-party dependencies, and interfaces exposed through APIs or user interactions. AIAST aligns with industry standards including OWASP LLM Top 10, OWASP ML Top 10, NIST AI RMF 1.0, ISO/IEC 42001 (2023) for AI Management Systems, and ISO/IEC 23894 (2023) for AI risk management. The framework provides deliverables including threat models, findings reports with CVSS scores, proof-of-concept exploits, risk registers, remediation plans, and compliance mapping documentation. The service targets industries including fintech, banking, cybersecurity automation tools, healthcare AI assistants, educational technology, retail chatbots, and government AI initiatives.