How is AI changing software engineering jobs?
AI is changing software engineering mostly by taking on routine coding and moving more of the engineer's day toward directing, checking and owning the work. The clearest warning sign so far is at the entry level, while the long-term picture for the whole profession is still mixed.
How developers are using it now
The 2026 Developer Survey from Stack Overflow drew more than 30,000 people. Among respondents who use AI coding assistants or agents, 73% use them daily. The most common development tasks are generating code in familiar areas (67%) and debugging (61%). Only 20% use AI for deploying, operating or troubleshooting production systems.
Keep in mind that people who answer a Stack Overflow survey choose to take it, so this is a picture of an engaged developer community and not of every engineer.
Trust comes with conditions. In the full 2026 Stack Overflow results, 48.0% of the 14,304 people who answered that question said they trust AI when they can easily verify the output. That matches what many engineers describe: the tool is useful, and you still have to read what it wrote.
What is happening to the work itself
The 2025 AI at Work report from Indeed Hiring Lab rated skills found in US job postings. It placed 81% of the skills in a typical software development posting in a group it calls hybrid. In that group, AI could do the bulk of the routine work while human oversight remains essential. Indeed describes this as potential, and says real outcomes depend on the tools, the business and the task.
So the skill that grows in value is judgment. Can you tell whether the code is right? Can you spot the security problem, the wrong assumption, or the feature nobody needed?
Does it really make engineers faster?
The evidence here is honestly mixed.
- The 2026 AI Index from Stanford says studies report gains of 26% in software development.
- A 2025 randomized trial from METR followed 16 experienced open source developers and found they took 19% longer with early 2025 AI tools, even though they believed they had gone faster.
- METR has since marked that result as out of date. Its February 2026 update says developers are likely sped up more in early 2026 than its early 2025 estimate showed, but calls its own new data very weak evidence of how much, partly because many developers no longer wanted to work without AI for a study.
Where the job market pressure is
The 2026 AI Index from Stanford reports that employment for software developers ages 22 to 25 has fallen nearly 20% from 2024, while most older age groups grew.
A related August 2026 update from the Stanford Digital Economy Lab, using ADP payroll data, found no widespread, economy-wide job displacement tied to AI. It did find young workers in highly exposed occupations about 19% below where they would be compared with peers in less exposed jobs, mainly through less hiring and not more layoffs. The authors say plainly that these are descriptive patterns and do not prove AI is the cause.
Other forces are in the mix too. Indeed Hiring Lab's June 2026 snapshot shows US software development postings at about 73 against a February 2020 level of 100, and it suggests AI-related roles may be helping a recent rebound. Meanwhile, the US Bureau of Labor Statistics projects employment of software developers, quality assurance analysts and testers to grow 10 percent from 2025 to 2035. That is a projection, and it can sit alongside a hard market for juniors today.
What I would do if I were job hunting
- Use the tools on real work. Build something, then be ready to explain what the AI got wrong and how you caught it.
- Get stronger at reading code. Reviewing, testing and debugging matter more when code arrives quickly.
- Learn how systems run. Few developers hand production work to AI so far, so operations, reliability and security knowledge can set you apart.
- Learn the business problem. Know why a feature exists and who it helps.
- Show proof. A small shipped project says more than a list of tools.
If you are early in your career, a crowded application pile is a hard place to stand out. A short note to the engineering manager lets a person see your work directly.
Here is a made-up example:
Hi Priya, I saw your team is hiring a backend engineer. I built a small payments retry service, used AI to draft parts of it, and found and fixed two race conditions in review. Could I send you the repo?
How Jobs Jetty helps
Jobs Jetty covers thousands of companies and their open roles in engineering, product, sales, marketing and recruiting. You can browse who's hiring by job and place to see what is open.
Upload your resume once. It shows you who to write to and drafts a note from your resume. Edit it so it sounds like you and points to your real work, then Send with Gmail opens your own Gmail with the note filled in, and you send it. Free is 10 notes a day, and sharing gets you more.
Try it at jobs.openjetty.ai.
FAQ
Will AI replace software engineers?
Nobody can state that as fact. Current research shows pressure on young workers in exposed jobs, no widespread displacement across the economy, and official projections that still show growth for developers.
Is computer science still worth studying?
It can be, especially if you learn fundamentals that help you judge and fix what AI produces. Look at real job posts in the area you want before you commit.
Should I mention AI tools on my resume?
Yes, if you use them for real work. Describe what you built and how you checked it instead of only listing tool names.