Software engineering is on the cusp of a profound shift with the rise of agentic artificial intelligence (AI). Gone are the days when AI was merely an assistant in coding and testing; now, it can autonomously manage entire projects. This evolution promises to streamline development processes but also requires significant organisational changes.

According to a recent survey of 300 engineering and technology executives, nearly half view agentic AI as their top investment priority for the coming year. While early gains are expected to be modest, teams anticipate accelerated delivery times, with an average increase in speed of 37% over two years.

The ultimate goal is full lifecycle management by AI agents within 18 months for many organisations, rising to 72% within two years. However, the transition won’t be seamless; compute costs and integration remain key challenges.

AI’s progression from mere tool to collaborator in software engineering signals a broader shift towards automation across industries. As it matures, agentic AI could transform how we build and manage software, but the journey ahead is fraught with complexities that need addressing.