Yes, putting company information into ChatGPT can be safe. It depends on which version you're using, what you type in, and whether anyone has actually set rules for it. The free, consumer version is the riskiest option. Anything typed into it can be used to train future models, unless that setting has been turned off. ChatGPT Enterprise and ChatGPT Team don't train on your data by default. That one difference explains most of the risk.
Most people ask "is ChatGPT safe" as if there's one ChatGPT. There isn't. The free version is built for individuals, and its default settings assume the person typing has nothing to hide. Company data isn't personal data. A client's name, a contract clause, a set of financial figures, these things belong to the business, not to whoever happened to be typing that afternoon.
The business versions come with a data processing agreement, admin controls, and training turned off by default. That's the version worth paying for once more than one person is using it for work. If your team is still on the free tier for anything client-facing, that's the first thing to fix, not the last.
Picture a project manager pasting a client's draft contract into ChatGPT to get a plain-English summary before a call. On the free tier, that contract has just left the building. Nobody decided that should happen. It happened because summarising a contract is genuinely useful, and nobody had told her not to.
That's the pattern behind most AI data leaks. It's rarely reckless. It's someone doing their job well, using the tool in front of them, with no guardrail telling them where the line is.
This worry isn't unusual. Research into Australian AI adoption has found privacy and security concerns to be the most commonly cited barrier to AI adoption, ahead of the equivalent figures in the US, UK and Canada. That's not businesses being behind. That's businesses being sensible about a question nobody has clearly answered for them yet.
The fix isn't avoiding AI. It's answering the question once, properly, so nobody has to guess.
A useful test: if you wouldn't paste it into a public forum, don't paste it into a free AI tool. Client names, financial data, anything under an NDA, anything a competitor would want, keep those out of consumer tools entirely.
Everyday writing help, brainstorming, summarising a public document, that's low risk on almost any platform. The line isn't about AI being dangerous. It's about knowing which tool you're holding and what it's built to do with what you give it.
A rule like that shouldn't live in one person's head. It belongs in a written AI policy the whole team can see, alongside a simple decision on when a human needs to check AI's work before it goes anywhere near a client. Australia's Voluntary AI Safety Standard sets out a similar approach at a national level, and it's worth reading once, even in outline (see our AI Standards overview).
None of this needs to be complicated. It needs to exist, and it needs to be something your team has actually read.
The businesses getting this right aren't the ones avoiding AI. They're the ones who took an afternoon to write the rule down, so their people could stop guessing and start using it properly. If AI workers are running inside your business rather than sitting in a browser tab, that decision gets made once, at the system level, instead of every time someone opens a new chat (see AI Workers).
Ready to see what AI transformation looks like for your business?
.svg)
.webp)