Is computer programming a stable career with AI taking over coding?

Computer Programmer scores 23/100 on AI exposure, one of the lower scores in tech, and BLS projects a 10% employment decline through 2033 as AI tools absorb routine code generation. The role is not disappearing overnight, but the 5,500 annual openings are shrinking and concentrated in organizations that still need programmers to maintain legacy systems and review AI-generated output. Programmers who expand into software development or DevOps are better positioned than those staying in traditional coding-to-spec work.

What does a computer programmer earn, and what moves the salary higher?

The median salary is $98,670, but programmers who specialize in high-demand languages like Rust, Go, or embedded C, or who transition into software development roles, typically reach $120,000 to $150,000. The biggest salary jump comes from moving beyond writing specifications into owning system design and debugging complex codebases. Programmers with strong code review skills, which are in demand as teams use AI to generate more code, also command a premium.

Which programming tasks is AI already handling, and which still need a human?

AI tools handle a meaningful share of routine code writing, template generation, and documentation, with automation rates near 30 to 44% in observed usage data. Troubleshooting across teams and debugging complex system interactions show high augmentation rates but low full automation, meaning humans are still in the loop. The work least affected is cross-team problem solving and translating ambiguous business requirements into technical specifications.

What skills help a computer programmer stay competitive as AI tools improve?

The most durable skill shift is from writing code to evaluating it: catching logic errors in AI-generated output, understanding system context, and debugging across large codebases. Proficiency in AI coding tools like GitHub Copilot is now a baseline expectation, not a differentiator. Programmers who pair coding ability with a domain specialty, such as financial systems, healthcare data, or embedded hardware, are harder to replace than those doing generic application work.

Risky • Entry-Level

Computer Programmer.

Create, modify, and test the code and scripts that allow computer applications to run. Work from specifications drawn up by software and web developers or other individuals. May develop and write computer programs to store, locate, and retrieve specific documents, data, and information.

Salary$98,670per year

Openings5,500per year

Growth-6%10 year

Example job titles

Analyst ProgrammerApplication Programmer AnalystComputer ProgrammerComputer Programmer AnalystInternet ProgrammerJava Developer, ProgrammerProgrammer AnalystWeb Applications ProgrammerWeb ProgrammerJunior Software DeveloperFull Stack DeveloperBackend DeveloperFrontend Developer

Why This Rating.

[01]

This role is risky

Computer programmers who primarily write code to specifications are among the most directly threatened by AI, as code generation tools can already produce working software from natural-language descriptions. The role lacks physical requirements, legal protections, or deep relationship-based value-it's defined almost entirely by producing a functional output that AI is rapidly learning to deliver. While experienced programmers who architect solutions and understand complex systems will adapt, the traditional role of translating requirements into code is shrinking fast.

Risks & Opportunities.

[02]

[01] Risks

-10%

employment expected to shrink over the next decade[1]

The core of this job, writing and maintaining code, already shows 33% automation in real AI usage data across over 14,000 observed conversations. Revising and expanding existing programs runs even higher at 36% automation. GitHub's 2024 research found that developers using Copilot completed tasks 55% faster, which means fewer programmer-hours needed per project [GitHub Blog, Sep 2024]. BLS projects computer programmer employment to decline 10% from 2023 to 2033, a loss of roughly 12,400 jobs, as companies consolidate programming work into broader software developer roles and offshore routine coding [U.S. Bureau of Labor Statistics, Sep 2024].

[02] Opportunities

17%

projected growth in software developer roles that programmers can transition into[1]

The strongest augmentation signal sits in consulting with teams to resolve problems in running programs: 77% augmentation rate across 747 conversations. AI acts as a diagnostic partner here, but the human still drives the fix. The 2024 Stack Overflow Developer Survey found that 76% of developers are using or planning to use AI tools, but only 43% trust the accuracy of the output [Stack Overflow, Jul 2024]. That trust gap is where experienced programmers add value: reviewing AI-generated code, catching subtle logic errors, and understanding system context that AI misses. Roles that combine programming skill with domain knowledge, like systems analysis or DevOps, are growing even as pure coding positions shrink.

Observed AI Impact.

[03]

AI is starting to automate these

Share of AI conversations where this task is handled solo, with little human involvement.

Test programs/applications

Conduct trial runs of programs and software applications to be sure they will produce the desired information and that the instructions are correct.

89.4% automated

Med usage

Write documentation

Compile and write documentation of program development and subsequent revisions, inserting comments in the coded instructions so others can understand the program.

77% automated

Med usage

Debug/correct errors

Correct errors by making appropriate changes and rechecking the program to ensure that the desired results are produced.

74.1% automated

High usage

Maintain software packages

Write, update, and maintain computer programs or software packages to handle specific jobs such as tracking inventory, storing or retrieving data, or controlling other equipment.

71.8% automated

High usage

Revise/expand programs

Perform or direct revision, repair, or expansion of existing programs to increase operating efficiency or adapt to new requirements.

66.6% automated

High usage

Write end-user manuals

Write or contribute to instructions or manuals to guide end users.

65.4% automated

Med usage

Write/analyze programs

Write, analyze, review, and rewrite programs, using workflow chart and diagram, and applying knowledge of computer capabilities, subject matter, and symbolic logic.

62.1% automated

High usage

Develop websites

Develop Web sites.

51.8% automated

High usage

Create workflow diagrams

Prepare detailed workflow charts and diagrams that describe input, output, and logical operation, and convert them into a series of instructions coded in a computer language.

51.2% automated

Med usage

AI hasn't figured these out

Low or no AI usage recorded for these tasks.

Consult with stakeholders

Consult with managerial, engineering, and technical personnel to clarify program intent, identify problems, and suggest changes.

No data

Data: Anthropic Economic Index (Jun 2026) · Real-world Claude usage patterns across occupations

How to Adapt.

[04]

[01] Already in this role?

Start using AI code-generation tools (GitHub Copilot, Claude, or similar) on real production tasks now, and track where they fail. The 2024 Stack Overflow survey found 43% of developers don't trust AI output accuracy [Stack Overflow, Jul 2024], which means the ability to audit, debug, and correct AI-generated code is immediately valuable. Over the next six months, build depth in areas AI handles worst for this role: cross-team troubleshooting, understanding legacy system context, and translating business requirements into technical specs. Those collaborative, context-heavy tasks show the highest augmentation rates and lowest automation rates in the data.

“The programmers who are thriving with AI coding tools are the ones who treat the output as a first draft from a junior engineer — always reviewed, often restructured. Speed went up. So did the importance of architecture judgment, because now you are reviewing a lot more code than you used to write.”
Gergely Orosz, The Pragmatic Engineer

“AI code generation has shifted the bottleneck in programming from writing code to understanding what you wrote. Programmers who can read, debug, and reason about large codebases quickly are more valuable than ever — that skill does not go away when the AI writes the first draft.”
Stack Overflow Developer Survey 2024

[02] Thinking of entering?

Pure coding skills alone are a shrinking market. GitHub's research shows AI is already handling a meaningful share of routine code generation [GitHub Blog, Sep 2024]. A stronger entry strategy: build a portfolio that demonstrates code review, debugging complex systems, and integrating components rather than writing greenfield code from scratch. Pair programming proficiency with a specific domain (healthcare data, financial systems, embedded hardware) to differentiate from both AI tools and generalist programmers. Consider aiming for software developer or DevOps roles rather than "computer programmer" titles specifically, since BLS projects software developers to grow 17% over the same period programmers decline [U.S. Bureau of Labor Statistics, Sep 2024].

“The entry bar has shifted from "can you write code" to "can you evaluate code, catch bugs in AI output, and reason about system design." Bootcamp graduates who understand that shift — and build portfolios that demonstrate it — are getting hired. Those who treat AI tools as a substitute for fundamentals are not.”
ThoughtWorks Technology Radar, Vol. 31

“67% of developers report using AI coding tools in their workflow, but the survey shows senior developers are adopting them at higher rates than juniors — not because they need more help, but because they can evaluate the output critically and redirect it when it is wrong.”
Stack Overflow Developer Survey 2024

The Career Map.

[05]

The full BLS career track, plus relevant emerging AI-native paths. Click any role to see how to transition to it.

Entry

Computer Programmer

★ This Role

Prompt Engineer

✦ Emerging

Mid-Level

Web Developer

15-1254.00

Risky

Web Designer

15-1255.00

Strong

Web Administrator

15-1299.01

Exposed

Software QA Analyst

15-1253.00

Risky

AI Integration Engineer

✦ Emerging

MLOps Engineer

✦ Emerging

Senior

Software Developer

15-1252.00

Strong

Database Architect

15-1243.00

Solid

Information Security Analyst

15-1212.00

Strong

AI Safety Engineer

✦ Emerging

Lead or Specialist

Computer Systems Architect

15-1299.08

Solid

Data Scientist

15-2051.00

Strong

IT Project Manager

15-1299.09

Solid

Information Security Engineer

15-1299.05

Strong

Principal

Computer Science Researcher

15-1221.00

Strong

Sources

0.Computer Programmer Occupational Data- O*NET Online

1.Occupational Outlook Handbook: Computer Programmers- U.S. Bureau of Labor Statistics, Sep 2024
2.Future of Work Report: AI at Work- LinkedIn Economic Graph, Jan 2024
3.Jobs on the Rise 2024- LinkedIn Economic Graph, Jan 2024
4.EU AI Act: first regulation on artificial intelligence- European Parliament, Aug 2024