How we use AI (and what we ask of you)
Our team are curious, polite and very human. We use AI to help us work better, never to replace the care we put into talking with you. Here is what you can expect from us, and the small things we ask in return.
What we promise when we use AI
AI is part of how we work now. It finds the right note in a long job history, summarises a forty-email thread and gets a first draft down quickly. None of that changes who is responsible for what we send you.
A person reads and checks everything
If AI helped draft a reply, a GravIT technician has still read it, checked it against your actual systems and put their name to it. You will never get an answer nobody here has looked at.
Short, specific and written for you
No padded emails and no generic copy-paste answers. If the fix is a restart, we will say so in a line. If it is complicated, we will explain it in plain English and tell you what it means for your staff.
Your data stays protected
We only use AI tools we have vetted, on business-grade plans where your information is not used to train the model. Passwords and MFA codes never go into an AI tool, and health, financial and children's information gets extra care.
Honest when we are not sure
AI can sound confident and still be wrong. When we do not know something yet, we will tell you, find out properly and come back to you. Ask us whether AI helped with anything we sent, and we will tell you that too.
What we ask of you
Using AI is fine. We do too. If writing does not come easily, or English is not your first language, AI can be a real help, and we would much rather hear from you than have you sit on a problem. These are requests, not rules.
- Read it before you send it. AI tools are good at sounding sure. Check the message still describes what is happening on your screen, because we will act on what it says.
- Tell us what actually happened. What you clicked, what you saw and any error message, ideally with a screenshot. Three lines in your own words often get you a faster fix than a page of polished description.
- Keep secrets out of public AI tools. Passwords, client records and anything we have sent you about your security settings are best kept out of free chatbots. If you are unsure whether a tool is safe for work, ask us.
- Be as kind to our team as they are to you. Everyone here is a real person doing their best to help, including on the day something has gone badly wrong.
If you would like to work with us
We read what people send us, so please skip the mass AI-written cold outreach and the automated follow-up sequences. Ask us a real question about our business and you will get a real answer from a real person, even if that answer is no thanks.
The same goes for quotes, proposals and support replies from the vendors we already use. If a person has checked it, we can rely on it, and that makes it much easier for us to recommend you to our clients.
Why we have written this down
AI makes it easy to send more words with less effort, and the cost of that lands on whoever has to read them. Good support depends on both sides saying clearly what they mean, so we would rather set the expectation once, kindly, than complain about it later.
It also helps with security. AI now writes very convincing scam emails, and a predictable, plain style from us makes it easier to spot a fake. We will never ask you to email us a password or MFA code. If something claiming to be from GravIT asks for one, ring us before you do anything.
Frequently asked questions
Does GravIT use AI to answer support requests?
AI helps us search our notes, summarise long email threads and draft replies faster. A technician reads every reply before it goes out, checks it against what is actually happening on your systems, and puts their name to it. If you would like to know whether AI helped with something we sent you, ask and we will tell you.
Is it alright to use ChatGPT or Copilot to write to us?
Yes. We use AI too, and if it helps you put a problem into words, use it. All we ask is that you read it before you send it, so it still describes what you are seeing. A short note in your own words with a screenshot is often the quickest way to get help.
Does AI see our business data?
Only the AI tools we have vetted, on business-grade plans where your data is not used to train the model. We do not put passwords or MFA codes into AI tools at all, and we take extra care with health, financial and children's information.
Will GravIT ever ask for my password by email?
No. We will never ask you to send a password or MFA code by email, whoever the email appears to come from. If something claiming to be us asks for one, or just reads oddly, ring 03 5280 8088 and check with a person.
Can you help us set expectations like this for our own business?
Yes. Our safe AI usage policies cover how your staff use AI tools, what data stays out of them and how AI output gets checked before it reaches a client.
Questions about how we use AI?
Ring the office and ask. A person will answer, and they will be glad you did.