Same tool, different rules
Welcome back! Last time we looked at what AI can and can't do in hospitality. Now let's ask a more practical question: which AI tool should you actually use at work, and on which account?
This matters more than it sounds. Think of it like the walk-in fridge. Two fridges can look the same. One is the venue's, it's checked and logged, and it's where the stock belongs. The other is a mini-fridge someone brought from home. You wouldn't keep the venue's stock in that one, would you?
AI tools work the same way. The app on your screen can look identical whether you're signed in with your personal email or your work account. But the rules about what happens to what you type can be very different.
Step one: find your workplace policy
So where do you start? With your workplace's rules, before you open any tool.
A lot of venues now have an AI policy. Sometimes it's a section in the IT or data protection policy, or in the staff handbook. Look on your staff app, the intranet or the shared drive, or ask in the office.
When you find it, look for answers to four questions. Which tools are approved? Which account should you use? What must never go into a tool? And who do you ask when you're not sure?
What if there's no policy at all? That's common, especially in independent pubs and cafés. It doesn't mean anything goes. Ask your manager, the owner, or whoever looks after IT or data protection. Something like: "I'd like to use an AI assistant to draft guest emails. Which tool and account are we OK with, and is there anything I shouldn't put in?" That one question protects you, your guests and the business.
Personal account or work account?
This is the big one. Let's look at the four tools we'll use as examples throughout the course. The settings and names change often, so always check the vendor's current help pages yourself.
ChatGPT. On personal plans like Free and Plus, OpenAI's Data Controls let you choose whether your chats help train its models. The setting is called "Improve the model for everyone". OpenAI says that on its business plans, ChatGPT Business and Enterprise, business data isn't used for training by default.
Gemini. In the personal Gemini app there's an activity setting called "Keep Activity". Google says some conversations may be read by trained human reviewers, and its privacy hub asks people not to enter confidential information they wouldn't want a reviewer to see. Gemini inside a work Google Workspace account is different. Google says Workspace data isn't used to train its wider models without permission.
Claude. On Anthropic's personal plans (Free, Pro and Max) there's a "Help improve Claude" setting that decides whether your chats can be used for training. Anthropic's business offering, Claude for Work, isn't covered by that consumer setting.
Microsoft 365 Copilot. If you sign in with your work account, Copilot Chat comes with what Microsoft calls enterprise data protection. Look for the shield icon at the top. Microsoft says that chat data isn't used to train the underlying models.
Do you see the pattern? Work accounts set up by your employer usually come with stronger protections and they're under the business's control. Personal accounts are under yours, and the defaults may be different. Even so, a work account doesn't make every kind of data OK to paste in. Your policy still decides that.
How to check a tool's settings yourself
You don't need to be technical for this. Here's a simple habit.
First, check which account you're signed in with. Your profile picture or email address usually shows it. Next, open the settings and look for anything about data, privacy or activity. Then look up the vendor's official help or privacy page, not a random blog. Last, if anything doesn't match your workplace policy, stop and ask.
Why the official page? Because these settings get renamed and changed, sometimes several times a year. What a colleague told you six months ago might be out of date.
Choosing the right tool for the job
So how do you pick? Start with what's approved, then match it to the task.
Say your hotel runs on Microsoft 365, so your email, rotas and documents live there. If Copilot is approved, it makes sense to draft a reply to a function enquiry with Copilot on your work account. It fits the systems you already use, and it has the protections the business has set up.
Now say your café uses Google Workspace and your manager has approved Gemini. Then that's your go-to for a first draft of a staff noticeboard message.
And what if your favourite tool isn't on the list? You might love the personal ChatGPT app on your phone. If it isn't approved for work, it isn't the right tool for work tasks, however handy it is. Instead, tell your manager why you'd find it useful. That's how policies get better.
When someone asks why you chose a tool, you should be able to say three things: it's approved, I used my work account, and it suits this task.
Key takeaways
- Find your workplace AI or IT policy first. If there isn't one, ask your manager, the owner or whoever handles IT or data.
- The same app can follow different data rules on a personal account and a work account. Check which one you're signed in with.
- Check data settings on the vendor's official help pages, because names and defaults change.
- A work account doesn't mean anything goes. Your policy still decides what you can share.
- Pick from the approved tools, then match the tool to the task and the systems your venue already uses.