Generally safe
These are good fits for AI. Read the output carefully before sending.
- Write a letter, email, or notice
- Draft a toolbox talk to refine
- Brainstorm ideas for a post or email
Prompt examples
Real prompts for real construction jobs. Tap any row to see the full prompt. Copy it, tweak it, paste it into your AI tool.
Featured and longer workflows
Use these when you need more than a one-line prompt. Start with the featured workflow, then copy and adapt to your project.
Build a Prompt
Answer a few questions and get a custom AI prompt for your task.
Step 1 of 5
What do you need to do?
Who is this for?
What format do you want?
Any extra details?
Your prompt result
Review for privacy before using any external tool.
Reuse your best prompts
A prompt is something you use for one task. If you regularly do the same type of work, save your instructions, examples, files, or background information in a Project or workspace so you do not have to start from scratch each time.
Use a prompt for a one-time task. Use a Project or saved workspace when you want to reuse the same instructions and context.
- Weekly site reports
- Proposal drafting
- Meeting summaries
- Safety communication
- Recurring client updates
Before you use any AI tool: read the prompt carefully. Remove any private info before you paste it in. Never share your cost breakdown or confidential project documents in a free AI tool.
Should I Use AI for This?
Figure out when AI is useful and when it might mislead you.
Is it right for this?
AI is great at some things, risky at others. Two quick gut-checks before you use it.
Quick check: what are you trying to do?
Match your task to a colour. If it's red, AI is the wrong tool. Use something else.
Useful, but verify
AI helps, but can miss or invent details. Always check the output against the source.
- Summarize notes, meetings, or documents
- Draft an RFI or change order
- Pull action items from a long email thread
Don't rely on AI
For these, AI is more likely to mislead than help. Use the proper source instead.
- Use AI to help you understand, not to confirm compliance
- Work out fixed pricing on a bid number
- Get advice on a contract or legal question
5 ground rules
These apply no matter what task you're doing.
Don't paste private project info into free AI tools.
Contracts, pricing, client details, signed documents. Free tools may use what you type to train future versions. Paid plans usually don't, but check the vendor's privacy page first.
Don't trust AI on regulations or legal advice.
AI can invent BC Building Code sections, WorkSafeBC rules, and contract clauses with full confidence. Always verify against the actual source before acting on it.
Check every important number
AI can make calculation mistakes or use the wrong information. Use a calculator, spreadsheet or approved estimating system, and check the inputs and result before relying on it.
Don't send AI-drafted emails without reading every word.
AI sometimes invents names, dates, prices, or commitments. Read carefully before anything goes to a client, owner, or inspector.
Don't use AI as the final word on safety.
A toolbox talk or safety briefing drafted by AI is a starting point. Your safety officer and WorkSafeBC requirements are the final word.
Spotted something else worth warning others about? Tell BCCA so we can add it to this list.
Flagship BCCA resource
Vet a vendor
Mark what the vendor can prove and what you still need to ask about. This is not a BCCA certification or score. It is a way to spot gaps before you buy.
0Evidence
found
found
0Need to
ask
ask
0Not yet
checked
checked
What should give you confidence
- Clear written answers and evidence you can verify
- References from real customers
- Transparent pricing and data-use explanations
- Clear support, implementation help, and data export options
Reasons to slow down
- Vague answers or claims that cannot be demonstrated
- Pressure to sign quickly or no customer references
- Unclear data use, ownership, or export options
- No explanation of what happens when the tool is wrong
If a vendor cannot clearly explain its product, pricing, data practices and support in writing, slow down and get those answers before you buy.
AI companies use big words. Here is what they actually mean.