Will AI Replace Software Engineers? What the Data Shows (2026)

We make Resymetry, a CV checker. We have no stake in how AI changes hiring beyond that, and we cite the page behind every number.

We looked for sources that measure something real: official job projections, job-posting data, payroll data, and published research on AI exposure and use. We preferred the newest edition we could read in full, and we read every page ourselves on 8 November 2026.

We quote each number as published and say what it measures. Most data is from the US, because that is where the detailed series exist. Where studies point in different directions, we say so instead of picking one.

We also cite our own count of 1,046 tech job postings, read on 6 November 2026. It shows what employers name, not what they strictly require.

The US Bureau of Labor Statistics (BLS) projects software developer jobs to grow 10% from 2025 to 2035. In 2025 there were about 1.7 million of them. BLS links part of that demand to building software for AI itself.

Job postings tell a rougher story. Indeed's index of US software development postings peaked in February 2022, at more than twice its February 2020 level. It fell to about 61 in May 2025. On 30 October 2026 it stood at about 82, still below where it was before the pandemic.

Most of that fall had happened by the end of 2023, when the index was about 73. That was before coding agents such as Claude Code, launched in February 2025. Since then, postings have risen. So falling postings alone do not show that AI is taking jobs.

Yale's Budget Lab tracks the whole US labour market for signs of AI. Its September 2026 update found no clear evidence of disruption linked to AI. It adds that this may change as new data arrives.

Source: US Bureau of Labor Statistics, Occupational Outlook Handbook: Software Developers, Quality Assurance Analysts, and Testers, read 8 November 2026; Software Development Job Postings on Indeed in the United States (FRED series IHLIDXUSTPSOFTDEVE), read 8 November 2026; Indeed Hiring Lab: AI and Job Postings: From Destruction to Creation? (July 2026), read 8 November 2026; The Budget Lab at Yale: Tracking the Impact of AI on the Labor Market (updated September 2026), read 8 November 2026

Research agrees that coding is among the work AI touches most. A 2023 OpenAI and University of Pennsylvania study found programming and writing skills linked to higher exposure to AI. Science and critical-thinking skills were linked to lower exposure.

Developers' own answers match. In the 2026 Stack Overflow survey, the top uses of AI are writing code, debugging and answering simple questions. Stack Overflow's write-up says only 20% use it for deploying, operating or troubleshooting live systems.

Using AI for a task is not the same as AI replacing the person. In a 2025 METR trial, 16 experienced open-source developers took 19% longer with AI tools, yet believed they had been 20% faster. METR now calls those results out of date, and says its newer data gives an unreliable signal. Productivity gains are still not settled.

Source: Eloundou, Manning, Mishkin and Rock: GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models, read 8 November 2026; Stack Overflow 2026 Developer Survey: AI, read 8 November 2026; Stack Overflow Blog: The results of the 2026 Developer Survey are here, read 8 November 2026; METR: Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity, read 8 November 2026; METR: We are Changing our Developer Productivity Experiment Design (February 2026), read 8 November 2026

The clearest split is between writing code to a spec and designing software. BLS projects computer programmer jobs, people who mainly write and test code others designed, to fall 7% from 2025 to 2035. It says this work keeps being automated, including with AI. Software developers, who plan and design systems, are projected to grow 10%.

Anthropic studied what people ask its Claude models to do with code. JavaScript and TypeScript made up 31% of queries, and HTML and CSS 28%. Building user interface components was the most common task, at 12%. That suggests web and front-end work is where AI is used most, among early adopters of one product.

At the other end, deploying and running live systems is where developers use AI least. No source we read measures job losses by specialty. So treat these as signs of where AI is used, not a list of jobs that will disappear.

• More exposed: routine coding to a clear spec, simple web and interface code, first drafts of tests.
• Less exposed so far: system design, running and fixing live systems, and work that needs judgement about trade-offs.

Source: US Bureau of Labor Statistics, Occupational Outlook Handbook: Computer Programmers, read 8 November 2026; US Bureau of Labor Statistics, Occupational Outlook Handbook: Software Developers, Quality Assurance Analysts, and Testers, read 8 November 2026; Anthropic Economic Index: AI's impact on software development (April 2025), read 8 November 2026; Stack Overflow Blog: The results of the 2026 Developer Survey are here, read 8 November 2026

This is where the evidence is most worrying. Stanford researchers used ADP payroll data on millions of US workers. For people aged 22 to 25 in AI-exposed jobs, employment is 19% below where it would be. The comparison is with peers in less-exposed jobs.

The gap comes mainly from fewer young people being hired, not more being let go. Experienced workers in the same jobs show no such gap. ADP's summary of the first version adds a software figure. In July 2025, employment of the youngest software developers was 20% below its late-2022 peak.

The authors call these early, descriptive signs, not proof that AI caused them. Some of the trends began before generative AI. Yale's tracker, which looks at the whole labour market, does not yet see a clear AI effect.

Employers also ask for more experience. From mid-2022 to mid-2025, the share of US tech postings asking for five years or more rose from 37% to 42%, Indeed found. In 2026, 71% of the recent rise in US software postings came from senior roles.

Source: Brynjolfsson, Chandar and Chen: Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence (Stanford Digital Economy Lab, revised August 2026), read 8 November 2026; ADP Research: Yes, AI is affecting employment. Here's the data. (August 2025), read 8 November 2026; The Budget Lab at Yale: Tracking the Impact of AI on the Labor Market (updated September 2026), read 8 November 2026; Indeed Hiring Lab: Experience Requirements Have Tightened Amid the Tech Hiring Freeze (July 2025), read 8 November 2026; Indeed Hiring Lab: AI and Job Postings: From Destruction to Creation? (July 2026), read 8 November 2026

AI is now part of many software job ads, though not most. In the 1,046 postings we read, large language models or generative AI were named in 46 of 217 software engineer postings. Machine learning was named in 44. AI coding tools were named in 15 of 82 full-stack postings.

The basics still lead. Python was named in 93 of the 217 software engineer postings. Monitoring and observability, seeing what a system does in production, was named in 84 of 214.

Indeed sees the same shift at scale. 37% of the rise in US software postings between May 2025 and May 2026 came from jobs with AI in the title. The full counts are in our post on 1,046 tech job postings.

Source: Resymetry job-postings counts (CSV), read 6 November 2026; Indeed Hiring Lab: AI and Job Postings: From Destruction to Creation? (July 2026), read 8 November 2026

Official projections still expect growth, and BLS lists a median annual wage of $135,980 for US software developers in May 2025. The UN's labour agency, the ILO, expects generative AI mostly to change jobs rather than remove them. Most jobs still need human input.

The harder part is getting in. Fewer entry-level hires and higher experience asks make the first job tougher than it was a few years ago. That is a reason to build evidence of real work early, not a reason to assume the field is closing.

Source: US Bureau of Labor Statistics, Occupational Outlook Handbook: Software Developers, Quality Assurance Analysts, and Testers, read 8 November 2026; International Labour Organization: Generative AI and Jobs: A Refined Global Index of Occupational Exposure (May 2025), read 8 November 2026; Brynjolfsson, Chandar and Chen: Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence (Stanford Digital Economy Lab, revised August 2026), read 8 November 2026; Indeed Hiring Lab: Experience Requirements Have Tightened Amid the Tech Hiring Freeze (July 2025), read 8 November 2026

If AI is changing the work, your CV should show the parts it does not do alone. That means decisions, systems you ran, and problems you found and fixed. Nothing here can promise an interview, but these changes make your evidence easier to see.

• Show design and judgement: what you chose, what you traded off, and why.
• Add production work: monitoring, on-call, incidents you resolved, systems you kept running.
• Name the AI tools and models you have actually used, and what you built with them. A bare "AI" line says little.
• If you are early in your career, lead with projects that someone uses, with links and results.
• Then read your CV against the actual job ad. Resymetry's free check does that and quotes the lines that match or are missing.

Source: Resymetry job-postings counts (CSV), read 6 November 2026; Stack Overflow Blog: The results of the 2026 Developer Survey are here, read 8 November 2026

• US Bureau of Labor Statistics, Occupational Outlook Handbook: Software Developers, Quality Assurance Analysts, and Testers — read 8 November 2026
• US Bureau of Labor Statistics, Occupational Outlook Handbook: Computer Programmers — read 8 November 2026
• Software Development Job Postings on Indeed in the United States (FRED series IHLIDXUSTPSOFTDEVE) — read 8 November 2026
• Indeed Hiring Lab: AI and Job Postings: From Destruction to Creation? (July 2026) — read 8 November 2026
• Indeed Hiring Lab: Experience Requirements Have Tightened Amid the Tech Hiring Freeze (July 2025) — read 8 November 2026
• Brynjolfsson, Chandar and Chen: Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence (Stanford Digital Economy Lab, revised August 2026) — read 8 November 2026
• ADP Research: Yes, AI is affecting employment. Here's the data. (August 2025) — read 8 November 2026
• The Budget Lab at Yale: Tracking the Impact of AI on the Labor Market (updated September 2026) — read 8 November 2026
• Eloundou, Manning, Mishkin and Rock: GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models — read 8 November 2026
• International Labour Organization: Generative AI and Jobs: A Refined Global Index of Occupational Exposure (May 2025) — read 8 November 2026
• Anthropic Economic Index: AI's impact on software development (April 2025) — read 8 November 2026
• Stack Overflow 2026 Developer Survey: AI — read 8 November 2026
• Stack Overflow Blog: The results of the 2026 Developer Survey are here — read 8 November 2026
• METR: Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity — read 8 November 2026
• METR: We are Changing our Developer Productivity Experiment Design (February 2026) — read 8 November 2026
• Resymetry job-postings counts (CSV) — read 6 November 2026