Test Generation Reloaded: How to Accelerate Test Design with AI & RAG — Without Breaking ISTQB Best Practices

Artificial Intelligence is making test case generation faster and more accessible, but speed alone does not guarantee quality, traceability, or maintainability. This paper explores how AI-powered assistants and Retrieval-Augmented Generation (RAG) can be applied to accelerate test design while remaining aligned with ISTQB best practices. It argues that effective AI-assisted test case creation depends not only on the model itself, but on the disciplined combination of context retrieval, prompt engineering, constraints, and output structure. The paper introduces a practical approach for using AI in test design by combining requirement and risk information with prompting techniques such as prompt chaining, few-shot prompting, and structured role-based instructions. It also discusses the limitations of large language models, particularly outdated public knowledge and the absence of internal organizational context, and presents RAG as a way to ground generated test cases in reliable project-specific information. In addition, the paper outlines a lightweight implementation path using local and privacy-aware tooling for experimentation and adoption. The session’s main contribution is a pragmatic framework for making AI a trusted partner in test case design rather than an opaque black box, enabling faster generation without compromising sound testing principles.