AI Tools introduced in Software Development. Analysis of Code quality, Security and Productivity Implications

Chatbots have recently been integrated into a diverse range of applications, from office tools to code compilers. It is undeniable that AI assistants possess features that simplify and automate repetitive tasks, however, it is yet unclear how well they can handle creative tasks such as writing code. This paper presents an in-depth analysis of the quality of the solutions generated by AI tools, that have been specifically trained for software development tasks, including code writing, testing, reviewing and documenting features. Individual use cases are rigorously examined, while comparing the results for two similar intelligent assistants per each scenario, to identify the strengths and weaknesses of their proposed solutions. The impact on productivity is studied, as it is dependent on the quality of the generated responses. The implementation of an incomplete, outdated or insecure solution will impose additional fixes, and, if the majority of answers have to be revised, this counterproductive behavior will actually increase the development time rather than facilitate its reduction.