Will AI replace programmers?
Updated
Programmers score 66 of 100 for AI exposure. The largest share of their task time, 75%, is work AI speeds up while they stay in charge. AI can take over 19% outright, and 6% stays with people.
- 19% AI can do it
- 75% AI helps
- 6% Needs a person
High exposure
Higher than 97% of the 923 occupations we scored.
Task by task
We graded 17 tasks, listed here from most to least exposed. AI can do 3 of them today, speeds up 13 and leaves 1 to people.
| Task | AI exposure |
|---|---|
| Compile and write documentation of program development and subsequent revisions, inserting comments in the coded instructions so others can understand the program. AI drafts program documentation and code comments; developer reviews accuracy. | 82% AI can do it |
| Write or contribute to instructions or manuals to guide end users. AI drafts end user manuals and guides from specs; writer edits. | 78% AI can do it |
| Write, analyze, review, and rewrite programs, using workflow chart and diagram, and applying knowledge of computer capabilities, subject matter, and symbolic logic. AI drafts and reviews code from flowcharts; programmer verifies logic and intent. | 72% AI helps |
| 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. AI converts flowchart descriptions into working code; programmer checks logic. | 70% AI helps |
| Correct errors by making appropriate changes and rechecking the program to ensure that the desired results are produced. AI spots likely bugs and suggests fixes; programmer confirms and retests. | 65% AI helps |
| Develop Web sites. AI generates site code and layouts quickly; developer is responsible for architecture and quality. | 65% AI helps |
| Perform or direct revision, repair, or expansion of existing programs to increase operating efficiency or adapt to new requirements. AI can refactor and modernize code; engineer validates behavior and performance gains. | 62% AI helps |
| 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. AI generates routine data and inventory code; developer is responsible for design and deployment. | 62% AI helps |
| Conduct trial runs of programs and software applications to be sure they will produce the desired information and that the instructions are correct. AI generates tests, executes them and reports results; person reviews failures. | 62% AI can do it |
| Perform systems analysis and programming tasks to maintain and control the use of computer systems software as a systems programmer. AI helps analyze configs and logs; privileged live system access still needed. | 45% AI helps |
Show the other 7 tasks
| Task | AI exposure |
|---|---|
| Investigate whether networks, workstations, the central processing unit of the system, or peripheral equipment are responding to a program's instructions. AI reads logs and suggests likely causes; hardware checks need engineers. | 35% AI helps |
| Consult with and assist computer operators or system analysts to define and resolve problems in running computer programs. Live troubleshooting with operators is interactive; AI helps read logs. | 33% AI helps |
| Train users on the use and function of computer programs. AI can draft training materials; in person user sessions need people. | 30% AI helps |
| Consult with managerial, engineering, and technical personnel to clarify program intent, identify problems, and suggest changes. Requirement clarification with stakeholders is interpersonal; AI only summarizes notes. | 25% AI helps |
| Train subordinates in programming and program coding. AI can draft course material; live teaching and mentoring stays with people. | 25% AI helps |
| Collaborate with computer manufacturers and other users to develop new programming methods. Collaborating with manufacturers and peers is relational; AI helps research methods. | 25% AI helps |
| Assign, coordinate, and review work and activities of programming personnel. Assigning and reviewing staff work is management judgment and accountability. | 15% Needs a person |
How programmers compare
All occupations (923)
Computer and mathematical (36)
Pay and outlook in the US
- Median pay
- $100,390 a year, May 2025
- People employed
- 92,230 in the US, May 2025
- Projected change in jobs
- -7.3% from 2024 to 2034
Skills to build
- Review AI-drafted code against flowcharts and specifications
- Deploy and roll back stacks in AWS CloudFormation
- Write Ansible playbooks that configure production servers
- Model data access patterns in Amazon DynamoDB
- Earn the AWS Certified Developer Associate certification
Safer roles nearby
5 related occupations with lower scores.
Questions about programmers and AI
Will programmers be replaced by AI?
Partly. About 5 in 10 working hours go to tasks AI can already do, mostly drafting program documentation and writing and reviewing code. What stays with people is the logic, the intent behind each program, and proof that the program produces the desired results.
Will AI take over coding jobs?
AI can already do 19% of the task time in coding and speeds up another 75%. People still do 6% of it. The US has 92,230 programmer jobs, with a projected change of -7.3% from 2024 to 2034.
Which programmer tasks is AI taking over?
AI drafts program documentation and code comments, the most exposed task at 82, because the code itself is the source material and the developer only checks accuracy. End-user manuals score 78, since AI can draft them from specs and a writer edits the result. Writing and reviewing code from flowcharts scores 72, with the programmer checking the logic.
Which programmer tasks stay human?
Assigning, coordinating and reviewing programming staff scores 15 on exposure, because management judgment and accountability belong to a person. Consulting with managers and engineers to clarify program intent scores 25, since requirement talks are interpersonal and AI only summarizes the notes. Live user training sessions score 30, because in-person sessions need people in the room.
How do programmers stay valuable as AI takes over tasks?
Move toward the work that still needs a person to finish, such as verification, deployment and stakeholder sessions. Review AI-drafted code against the flowchart and retest every fix. Take on deployments in AWS CloudFormation and Ansible, and lead requirement sessions before coding starts.
Related jobs
Next steps for programmers
Check which roles in your team AI can take on
We map your team's tasks, show what AI can do today and build the automations that pay back first.
Related topics
- Jobs AI can't replaceCement masons score lowest, at 8.
- Jobs most at risk from AIProofreaders score highest, at 76.
- High-paying jobs with low AI exposureRadiologists earn a median $420,860 and score 46.
- Entry-level jobs with low AI exposureCement masons need little prior training and score 8.
- Jobs AI speeds up the mostFor natural sciences managers, AI speeds up 96% of task time.
- AI and jobs in Malta15 iGaming roles graded task by task.
- Skills that keep work humanThe skills to build in the jobs AI changes least.
- Jobs A to ZEvery job we publish, by first letter.
- AI exposure by industryAverage scores for 22 occupation groups.
- Write a CV screening software can readHow applicant tracking systems read a CV, and what to change.
- AI for businessesThe tasks to automate first in each team and industry.
- How we score AI exposureThe grading, the weights and every source.