Most of the AI news in construction is about robots, drones and self-driving equipment. That's not what most contractors are doing with it.
Walk into a contractor's office today and you'll find AI helping with the paperwork. A project manager uses it to write an RFI. An estimator uses it to pull the key points out of a 400-page spec. A payroll admin uses it to find out who still hasn't turned in their time.
Most of this happens in general tools like ChatGPT, Claude, Copilot and Gemini. Some of it happens in software built for construction, like estimating, photo and payroll tools. Either way, the people using it still check the work before it goes out. A wrong quantity or a missed contract clause costs too much to skip that step.
So how many contractors use AI? The survey numbers are all over the map. Bluebeam surveyed 1,000 AEC professionals and found 27% use AI in their operations. RSM's 2025 survey found 95% of its construction respondents use generative AI, though only 80 construction firms took part. The two surveys asked different groups of people, which explains some of the gap.
What's clear is that more contractors are trying it every year. Below are 10 ways they're using AI right now, from the office to the jobsite, along with the spots where you'll want a person to double-check.
Vyki finds missing time, chases approvals and checks overtime before payroll.
| If your team spends hours on... | Try AI for... |
|---|---|
| Emails and RFIs | First drafts |
| Long specs and contracts | Summaries and key-term lists |
| Meeting notes | Clean summaries and action items |
| Safety training | Toolbox talk drafts |
| Timesheets before payroll | Missing time and overtime checks |
AI makes mistakes, and it doesn't always flag them. It can misread a table or quote a spec section that isn't there. In construction, a bad quantity or a wrong pay rate costs real money. Treat what AI gives you as a draft, and have the person who owns the work review it.
The top challenges RSM's construction respondents reported were data quality (36%), budget (32%) and data privacy and security (29%). Before you paste in a contract or employee data, know where that data goes.
We built Vyki, our AI payroll assistant, for number 10 on this list. She works inside hh2 Time Tracking to find missing time, send reminders, chase approvals and check timesheets against federal and state overtime rules. Your Payroll Manager confirms changes and gets a summary before each payroll run.
Contractors use AI to draft emails and RFIs, sum up specs and contracts, help with estimates and takeoffs, sort site photos, draft safety training, translate for crews, write daily reports and prep payroll. Most start with general tools like ChatGPT, then add construction-specific tools for estimating, photos or payroll.
Surveys vary. Bluebeam's survey of 1,000 AEC professionals found 27% use AI in their operations. RSM's 2025 survey found 95% of its 80 construction respondents use generative AI. The gap likely reflects who each survey asked.
In RSM's 2025 survey, construction respondents using generative AI most often named ChatGPT (83%), Microsoft Copilot (51%) and Google Gemini (43%). Many contractors also use AI built into construction software for estimating, photos, scheduling or payroll.
It can be, with rules. Don't paste in sensitive data like Social Security numbers or bank details. Check anything that affects cost, safety or pay. And set a simple company AI policy so everyone knows what's allowed.
If you're just getting started, pick one task that eats up your team's week. Try AI on it for a couple of pay periods and see if it helps.
If that task is timesheets before payroll, book a demo and see how Vyki handles it.