By Lee Gilmore, AWS Practice Lead at Leighton
Artificial intelligence is fundamentally reshaping how organisations operate, compete, and create value, transforming everything from decision-making and customer experience to product innovation and workforce operations.
In software engineering AI is changing how digital products are designed, built, and delivered today.
Some organisations are embracing AI rapidly and seeing major productivity gains. Others remain cautious, questioning the reliability of these systems and the risks of overdependence. Both perspectives are valid. AI is incredibly capable, but it is not infallible. It can accelerate innovation just as easily as it can accelerate technical debt without the right oversight.
That is why the conversation around AI in software engineering should not focus on replacement. It should focus on elevation.
The role of the engineer is shifting from writing every line of code manually to orchestrating increasingly capable AI systems. The value is no longer in typing faster, but in thinking more clearly. Engineers who understand architecture, systems design and business context are becoming exponentially more effective when paired with AI-enabled tooling.
This mirrors what happened in industries such as radiology, where AI augmented specialist practitioners rather than replacing them. Repetitive work became automated, allowing professionals to focus on judgement, interpretation and complexity.
As the cost of generating code rapidly approaches zero, the real differentiator becomes the ability to generate the right code in the right way. This is creating demand for a new kind of engineering leader: the architect developer.
The architect-developer sits between traditional architecture and hands-on software engineering. They understand cloud platforms, engineering standards, product thinking and AI tooling. Their role is to guide AI systems, validate outputs and ensure that speed does not come at the expense of maintainability, governance or business alignment.
This evolution is also changing how teams operate. Traditional large-scale delivery models were built around the assumption that humans manually produced all software. AI changes that equation. Smaller, highly skilled teams can now deliver outcomes at a pace that previously required significantly larger departments.
At Leighton, we are increasingly seeing autonomous “pods” of engineers, guided by architect-developers and supported by AI agents, delivering software in dramatically shorter cycles. In many cases, allowing working software to be generated and iterated in hours rather than weeks.
However, speed alone is not the goal. Governance and context are becoming more important than ever.
Organisations succeeding with AI are not simply adopting tools quickly; they are building the frameworks, standards and architectural guardrails that allow AI to operate responsibly. In many ways, an organisation’s most valuable intellectual property is no longer just its codebase, but the engineering principles, domain knowledge and operational context that shape how AI generates solutions.
The next phase will go even further. We are moving towards generative-native systems where AI assists not only with writing software, but also with monitoring, remediation and optimisation across entire platforms.
For businesses, the message is clear. AI will not eliminate the need for engineers, but it will redefine what great engineering looks like. The organisations that thrive will be those that invest in adaptable teams, modern platforms and people capable of orchestrating AI with purpose, judgement and accountability.
At Leighton we’re working with our customers to help them better understand how they can integrate AI into their teams and workflows through our AI-driven workshop. If you’re interested in exploring how you can better enable your teams to deliver in an AI era we’d love to hear from you: hello@leighton.com