Software engineers were among the first to work alongside AI every day, through coding assistants and now coding agents. That makes the question personal for many of them: if AI can write code, what is left for the people who used to write it? The data gives a split answer, and the split itself is the most useful part.
The short answer: software development is significantly exposed to generative AI, and exposure has been rising. But the evidence points to a shift in the job rather than its disappearance: US projections show software developer employment growing while computer programmer employment declines, with AI named as a reason for the decline. Writing routine code is moving to AI; designing systems, deciding what to build and checking the result are moving to the center of the job. Forecasts can change quickly in this field.
How exposed software jobs are
Scores below come from the ILO's 2025 index: each task is rated from 0 (generative AI cannot do it) to 1 (it could be done without a human), and occupations are banded from not exposed and minimal exposure up to gradients 1 to 4. Our overview explains how to read them.
| Occupation (ISCO-08) | ILO category | Mean task score (0 to 1) |
|---|---|---|
| Web and Multimedia Developers | Gradient 4 (highest) | 0.60 |
| Applications Programmers | Gradient 3 | 0.57 |
| Database Designers and Administrators | Gradient 3 | 0.57 |
| Software Developers | Gradient 3 | 0.53 |
| Web Technicians | Gradient 3 | 0.53 |
| Systems Analysts | Gradient 2 | 0.49 |
Between 2023 and 2025 the ILO raised its scores for several digital occupations, including web and media developers and software-related jobs. It ties this to AI's new agentic abilities: carrying out multi-step tasks on its own and working inside software environments, with software development among the leading adopters. It also notes that these occupations have a strong capacity to evolve and develop new tasks using AI.
Developers up, programmers down
The US Bureau of Labor Statistics separates the two:
- Software developers, quality assurance analysts and testers: employment projected to grow 10 percent from 2025 to 2035, with about 106,100 openings a year on average (BLS Occupational Outlook Handbook).
- Computer programmers: employment projected to decline 7 percent over the same period. The BLS explains that programming work continues to be automated and that many companies are expected to use technologies including AI to automate repetitive programming tasks, while some higher-skilled programming tasks will likely shift to other workers, such as software developers (BLS Occupational Outlook Handbook).
The WEF's 2025 employer survey also lists software and applications developers among the fastest-growing roles in percentage terms (WEF Future of Jobs Report 2025).
What AI is taking on
- Boilerplate, routine functions and standard patterns.
- Test scaffolding and simple bug fixes.
- Explaining unfamiliar code and drafting documentation.
- Translating code between languages and frameworks.
- Multi-step changes carried out by coding agents under supervision.
What moves to the center
- Deciding what to build: turning a vague need into a clear specification.
- System design: architecture, trade-offs and how pieces fit together.
- Reviewing AI-written code: correctness, edge cases, maintainability and security.
- Security: agents that can run commands and change code bring new risks, such as prompt injection and excessive agency (OWASP Top 10 for LLM Applications 2025).
- Owning production: someone still answers when the system fails at 3 a.m.
How the job is changing
For many engineers, the day is shifting from writing every line to briefing agents, reviewing their changes and integrating the results: directing and checking the work rather than doing every step. Entry routes may be the most affected, since routine coding tasks were how many juniors learned. Engineers who combine solid fundamentals with strong review skills and domain knowledge are likeliest to do well.
Read next: AI agent security, checking AI output and AI agents as coworkers.
Pay and how to get in
The US Bureau of Labor Statistics puts the median annual pay for software developers at $134,040 in May 2025, against $50,980 for all workers. Across the 2 BLS occupations this page covers, median pay ranges from $100,390 to $134,040. The typical education needed to enter is: bachelor's degree. See how this job compares in AI exposure by pay and AI exposure by education level.
Questions and answers
Will AI replace software engineers?
Current evidence points to a shift rather than replacement. The BLS projects 10 percent growth for software developers from 2025 to 2035, while programmer employment is projected to fall 7 percent, partly due to AI automating repetitive programming.
How exposed are software developers to AI?
The ILO places software developers in gradient 3 (significant exposure) and web and multimedia developers in gradient 4, and raised scores for several digital jobs between 2023 and 2025.
What is the difference between programmers and developers in the projections?
The BLS expects repetitive programming tasks to be automated and some higher-skilled programming work to shift to software developers, whose employment is projected to grow.
What skills matter most for engineers working with AI?
Specification, system design, reviewing AI-written code, security awareness and owning production systems.
Is software engineering still worth going into?
Projections still show growth for developers, but the work is changing. Strong fundamentals plus the ability to direct and review AI agents are increasingly central.