Microsoft Copilot executive reveals how AI is changing software engineering

Microsoft addressed hundreds of security vulnerabilities in its products during last month’s Patch Tuesday, and the company is expected to release similar updates this month as it continues to integrate AI into its development process.

Although there is some skepticism about this approach, AI is now a central part of enterprise software development. A Copilot executive recently outlined how AI is changing the way Microsoft builds its software.

Charles Lamanna, a longtime Microsoft leader, stepped into the role of Executive Vice President of Copilot, Agents, and Platform six months ago after a major reorganization at the company.

In a recent post on X, Lamanna described how Microsoft is weaving AI into its development work, while preserving the same standards for customer success.

Lamanna noted that his team’s biggest accomplishment this year is not tied to launching a major product or feature.

Instead, he pointed to a fundamental change in the team’s approach to software development, using agentic coding and other AI tools.

He explained that this new way of working is not limited to coding or his own department. Teams that fully embrace AI in their development processes frequently see better results than those that only add AI on the surface.

Engineers are now guiding agentic systems to handle programming tasks, which allows them to spend more time on strategic decisions, system architecture, and quality control.

With coding less of a barrier, the main challenge has become making the right choices to deliver the best possible product.

READ MORE: Microsoft denies Windows 11 background service tracks users without consent

Lamanna says that even though AI has enabled engineers to build many more features by freeing up time, he is guiding his team to build the same number of features while using that surplus time to focus on reliability, performance, and security.

One result of this new approach has been a considerable drop in Power Automate support tickets compared to last year. Lamanna emphasized that customer success metrics and KPIs should stay the same, even as AI becomes more integrated.

Lamanna also stressed that engineering teams should not add AI just for the sake of it, but should rethink how they work with the technology at a fundamental level.

Teams that succeed in this approach can show the value of adopting AI and encourage others to do the same. This strategy has helped Microsoft move products from idea to market in just a few months, something that would have been hard to imagine only a few years ago.

Lamanna admitted that Microsoft still meets challenges. This is not surprising, as there have been reports that the company is struggling to fix security bugs as fast as AI can find them.