JPMorgan Chase CIO Gill Haus is redefining what makes a software engineer valuable to the bank. Generative AI tools are automating more of the coding process, shifting the focus from writing code to knowing what code to write. The change affects not just who gets hired, but how development teams operate and what skills matter most.
"We don't really hire engineers to write code -- we hire them to know what code to write," Haus said. "They must understand the problem and use technology to solve that problem." Before AI, engineers spent significant time on manual coding, configuring environments, and writing tests. Now, those tasks can be automated, freeing engineers to concentrate on customer needs and system design.
Engineering fundamentals matter more
Haus stressed that AI does not diminish the need for core software engineering skills. Instead, it amplifies them. Automated code generation requires rigorous testing to catch errors that a human might not have made. "If you have a computer now writing code for you, there's a ton of testing that needs to be done," he said. "We can't keep up with that unless we automate it."
As a result, practices like automated testing, continuous deployment, and automated rollback become essential controls. Good architecture, security, and governance are also critical. AI tools can recommend architecture, but an engineer still needs the judgment to decide if that recommendation fits the system's requirements. The fundamentals taught in computer science degrees -- testing, design, security -- remain relevant, but they now need to be applied at a faster pace.
The expanding role of non-engineers and the need for specialists
AI tools will let non-engineers build more software, working alongside engineers to solve business problems. However, Haus drew a clear line. "If you had a million customers on that, how would it work if there was a failure in one of your data centers? How do you handle security? What do you do if somebody wants to delete their account?" Those scenarios demand deep system understanding that only experienced engineers possess.
This means teams will likely blend technical and product skills, but specialists remain necessary for scalability, reliability, and secure system design. "Code has transformed. Code is becoming English," Haus said, noting that the real value now lies in tackling the backlog of features that were previously deprioritized because of time constraints.
As code generation becomes automated, many IT professionals are exploring Generative Code courses to understand how AI-assisted coding integrates into the software delivery lifecycle.
Managing risk in regulated environments
At a highly regulated institution like JPMorgan Chase, security and privacy are paramount. Haus said the bank keeps a human in the loop for any customer-facing AI output and focuses on replacing manual processes with automated controls. "Confidence in production comes from strong engineering practices -- automated testing, automated deployment and automated rollback," he said. With those controls, AI agents writing code can be monitored and rolled back quickly if issues arise.
He noted that adoption inside the organization is mixed. Some engineers are excited and eager, while others are anxious about job security. Haus emphasized transparency and communication, telling teams that change is coming but the company will support them. "I understand why people get scared of AI, but employees are the ones who still have the agency to decide what agents should do," he said.
The technology is still maturing. "It's only been a few years since ChatGPT became a big thing. Claude Code didn't reach its current level of capability until late last year, so we're all still learning what this means for how people will work," Haus said.
Why this matters for IT and development
The message for IT and development professionals is clear: coding ability alone is no longer the primary currency. Understanding customer problems, exercising judgment about what to build, and mastering engineering fundamentals are what will separate effective contributors from the rest. Automated testing, deployment, and rollback are no longer nice-to-haves; they are prerequisites for safely integrating AI into the development pipeline. For those looking to adapt, AI for IT & Development courses offer a path to building the skills that matter in an AI-augmented workflow.
Your membership also unlocks:
