Will AI replace Plumbers, Pipefitters, and Steamfitters?
AI reaches little of this job’s work.
5/100
Low exposure
More exposed to AI than 11% of US workers.
- AI already used in
- 4% of the work
- As AI tools mature: up to
- 10% of the work
What the job involves
Assemble, install, alter, and repair pipelines or pipe systems that carry water, steam, air, or other liquids or gases. May install heating and cooling equipment and mechanical control systems. Includes sprinkler fitters.
Where AI reaches the work
The job’s main tasks, heaviest first, by how far AI reaches them.
Already happening(1)
- Review blueprints, building codes, or specifications to determine work details or procedures.
Within reach(3)
- Lay out full scale drawings of pipe systems, supports, or related equipment, according to blueprints.
- Keep records of work assignments.
- Estimate time, material, or labor costs for use in project plans.
Out of reach(26)
- Assemble pipe sections, tubing, or fittings, using couplings, clamps, screws, bolts, cement, plastic solvent, caulking, or soldering, brazing, or welding equipment.
- Locate and mark the position of pipe installations, connections, passage holes, or fixtures in structures, using measuring instruments such as rulers or levels.
- Cut, thread, or hammer pipes to specifications, using tools such as saws, cutting torches, pipe threaders, or pipe benders.
- Install underground storm, sanitary, or water piping systems, extending piping as needed to connect fixtures and plumbing.
- Fill pipes or plumbing fixtures with water or air and observe pressure gauges to detect and locate leaks.
and 21 more
Also called
Pipefitter · Plumber · Sprinkler Fitter · Plumbing Installer · Steamfitter · Commercial Plumber · Fire Sprinkler Installer · Gas Main Fitter
Related jobs
What these numbers mean
The score is the share of this job’s work, weighted by how important and frequent each task is, that AI could do in half the time: tasks an AI model could do alone count in full, tasks that need software built on AI count half.
“AI already used” is the share of the work in tasks where people are recorded using AI today. “As AI tools mature” counts every task AI could halve, with or without extra software.
The percentile compares the score with every job’s, weighted by how many people in the US do each one. Exposure means AI could change how the work is done, not that the job goes away. These are estimates for the typical job, not predictions.
Sources
- O*NET 31.0 Database, U.S. Department of Labor, Employment and Training Administration
- GPTs are GPTs: Labor market impact potential of LLMs (task labels), Eloundou, Manning, Mishkin & Rock (2023), OpenAI and University of Pennsylvania
- Anthropic Economic Index: labor market impacts (task penetration), Anthropic
- Occupational Employment and Wage Statistics, May 2025 (national and metropolitan areas), U.S. Bureau of Labor Statistics