Generative AI models that generate source code or structured data are becoming important tools in software engineering. However, erroneous or incorrect outputs can cause faults leading to errors in software applications. In this workshop we will discuss these risks and approaches to solve them. In particular, neurosymbolic methods that combine data-driven learning systems and symbolic, logical approaches represent a promising path.

Submissions will centre on the following themes, while also considering additional relevant topics: automatic code generation, aiops, generative ai in software engineering and trustworthy ai in software engineering.