Fiona Fung, Engineering Director for Claude Code and Claude Cowork, introduced the concept of AI-native engineering organizations, where the primary focus shifts from writing code to its verification, auditing, and security assurance.
What Happened
Fiona Fung described the transformation of the software development lifecycle. In new processes, AI agents take over routine tasks such as style compliance, bug fixing, and writing tests. Human specialists are transitioning to critical functions: managing legal risks, ensuring security, and shaping "product taste."
Context
Tools like Claude Code and Claude Cowork are beginning to implement new standards where AI agents act not just as autocomplete tools, but as full participants in the development process. This leads to a profound transformation of engineering culture, where AI agents become a standard element of the CI/CD pipeline.
Why It Matters for the Industry
For the industry, this means a shift in efficiency metrics. Instead of the volume of code written, the focus moves to onboarding speed, PR cycle time, and the percentage of commits performed with the help of AI. Roles are being redefined: managers are moving closer to technical implementation, while engineers are moving closer to system design and architecture.
Why It Matters for Users
For developers, writing code is becoming a "commodity," while architecture, verification, and compliance resolution are becoming the most valuable and expensive skills. To remain relevant, specialists must develop systems thinking and the ability to conduct deep audits of code created by AI agents.
What Is Not Yet Known / Limitations
Legal representatives and enterprise architects highlight specific risks in IP, compliance, and security as independent critical factors, whereas other roles view them more as part of the general system design process.
Sources
Author
Look at AI, Editorial Team