AI in Cost Engineering: Transitioning from Curiosity to Daily Use

SPCEA recently ran a multiple-choice poll on LinkedIn asking a simple question:
“How are you MOSTLY using AI in your current role?”

The responses offer useful insight into where cost engineers are right now with AI. Not where the hype says we are. Not where vendors say we should be. Just where people actually seem to be using it in their work.

Here’s what the poll showed . . .

The survey results are rounded, so the total responses do not add exactly to 100%. Nevertheless, the big takeaway is hard to miss: cost engineers are actively using AI as part of their jobs.

Only 3% of respondents said they are not using AI at all. That means the overwhelming majority are already putting AI to work in some form. For a field like cost engineering, where accuracy, judgment, documentation, and traceability matter a lot, that’s a pretty important shift.

Most are Using AI as a Knowledge Assistant

The largest share of respondents, 49%, said they are mostly using AI for knowledge assistance.

For many cost engineers, AI is becoming a faster way to perform basic office tasks. For example: drafting documents, organizing notes, checking assumptions, explaining unfamiliar topics, or getting a first-pass structure for a report or presentation. It is also being used to research manufacturing processes and learn the cost drivers behind them. That can be especially helpful when an engineer is looking at an unfamiliar part, material, process, or supplier quote and needs to quickly understand the cost drivers.

That kind of use is practical. It does not require a major software build or formal programming skills. It just requires someone willing to try the tool on real work and see where it saves time.

Of course, AI-generated research still needs to be checked. Manufacturing methods, labor requirements, yields, cycle times, materials, tolerances, tooling, inspection needs, and production volumes can all affect cost in ways that are very specific to the job. AI can help point the engineer in the right direction, but the final judgment still belongs to the cost professional.

Nearly as Many Cost Engineers are Building Tools

The second result is just as interesting: 46% said they are building AI tools for their teams.

That is a strong signal, as it suggests that cost engineers are not only using off-the-shelf AI tools but also shaping them around internal workflows. This could include custom estimating agents, proposal support tools, historical data search tools, spreadsheet “helpers”, or systems that help teams find and reuse historical project information.

That matters because cost engineering work is often highly specific. Generic AI can be helpful, but the real value usually comes when specific agentic AI understands the organization’s format templates, cost structures, terminology, historical data, and review process.

Still, this area comes with responsibility. Building AI tools for a team means thinking carefully about data quality, permissions, validation, and how outputs will be checked. A tool that gives a fast answer is useful only if the team understands where that answer came from and what still needs expert review.

Very Few Are Avoiding AI Entirely

The “not using it” response came in at only 3%. This indicates that complete non-use may already be the exception rather than the norm among the survey respondents.

For cost engineers who have not started yet, the easiest entry point is probably not automation - it is assistance. Use AI to summarize something, rewrite something, organize a set of notes, research an unfamiliar manufacturing process, or create a checklist. Start with low-risk tasks. Then build from there.

What This Means for Cost Engineers

Cost engineers appear to be using AI in two main ways: first, as a productivity and knowledge aid, and second, as a foundation for team-level tools. Both paths are important. One improves the way individuals work, while the other may eventually change how estimating teams manage knowledge, analyze data, and deliver products.

The next challenge is not simply to " use AI more.” It is using AI well. That means keeping the cost engineer in control. It means checking outputs and protecting sensitive information. It means being clear about assumptions. And, most importantly, it’s remembering that AI can help with speed, structure, research, and search, but it does not replace professional judgment.

Jeff Miller Jeff Miller is President and Co-Founder of SPCEA and has 40 years of engineering, manufacturing, and commercial experience within the electronics and semiconductor industries. He has served in leadership and direct-contributor roles at General Motors, John Deere, Standard Motor Products, Ford Motor Company, Whirlpool Corporation, and Panasonic Automotive Systems. Jeff has been active within the cost engineering profession since 2002.   Copyright © 2026 – Society of Product Cost Engineering & Analytics. All rights reserved