As AI rapidly automates coding tasks on software development sites, the capabilities demanded of developers are also changing. In addition to the ability to write code directly, judgment and communication skills are becoming more important, including verifying AI-generated outputs and reflecting customer requirements in product design.
Business Insider reported on Nov. 16 that U.S. technology companies are demanding a broad range of roles from developers, from managing AI coding agents and reviewing code to collaborating with customers.
Michael Irwin (마이클 어윈), a senior software engineer at Docker, recently showcased a wooden game controller he made at a conference. Using a device connected to a game written by AI, Irwin discussed with visitors how to use AI agents. "Thanks to AI, I can use the time I used to spend on detailed coding for other activities," Irwin said, adding that developers need a new way of thinking.
Companies' development methods are also changing. Amazon Web Services recently wrote the entire source code of its newly introduced local testing tool, the AWS Emulator, using autonomous AI agents. It applies the code through automated testing and verification systems.
The range of jobs using AI coding tools has also expanded. OpenAI said that about 20 percent of Codex users in June were in roles other than developers. As a result, developers are being required not only to have expertise to judge errors and security issues in AI-made code, but also the ability to communicate with non-technical departments.
An internal survey Anthropic disclosed last month showed meeting time for technical practitioners rose 16 percent from a year earlier. As the ratio of engineers to product managers climbed to about 17 to 1, a trend has emerged of developers directly participating in product-related decision-making.
Hiring methods are also changing. The share of technology job postings tallied by U.S. staffing firm Robert Half that require office attendance rose to 84 percent this fall from 63 percent in early 2024.
A CompTIA survey also showed the share of software engineer job postings that specify communication skills rose to about 60 percent recently from 47 percent in early 2020. Still, it is not possible to conclude that the expansion of office attendance is solely the result of AI adoption. Changes in the hiring market and companies' return-to-office policies may also have had an impact.
It is also difficult to say that developer jobs themselves have sharply declined. The U.S. Bureau of Labor Statistics counted about 1,900,000 people in software development, quality assurance and testing in 2025, and projected related employment to rise 10 percent by 2035.
By contrast, barriers to entry for junior developers could rise. In Anthropic's new hiring for technical roles, the share of junior hires fell to 9 percent in May from 27 percent in March last year. Concerns have also emerged that opportunities to build experience could shrink as AI performs entry-level coding tasks.
Analysts say developers' competitiveness in the AI era depends not only on the ability to generate code quickly, but also on technical judgment to verify the accuracy of outputs, solve complex problems and collaborate.