Your staff's AI chats may be kept for three years
Google's Gemini privacy page says some chats are read by people and kept for up to three years. What that means when staff use personal AI accounts.
Abhijit Chatterjee Chattopadhyay

Did someone on your team paste a customer email into a free AI app this week, to help write the reply? Google's own privacy page for its Gemini app explains where a chat like that can end up: in front of a human reviewer, and kept for up to three years.
This is not a story about Google doing something wrong. The page is clear and the settings exist. The risk is that nobody in your business has read it.
What Google's page says
The Gemini Apps Privacy Hub (opens in a new tab), last updated on 26 October, says that "Human reviewers (including trained reviewers from our service providers) review some of the data we collect" to improve Google's services.
It also says: "Chats reviewed by human reviewers (and related data like your language, device type, location info, or feedback) are not deleted when you delete your activity. Instead, they are retained for up to three years."
There is a setting that stops future chats from being reviewed to improve Google's services, and a temporary chat mode. Neither closes the door completely: the page says that even with the setting off, or in a temporary chat, Google still uses chats to protect Google, its users and the public, with help from human reviewers. The page also says that "If you have a work or school Google Account, your use of Gemini Apps may be subject to different data handling terms."
So the same app can handle the same text very differently, depending on which account your colleague was signed in to.
Why this is your problem, not theirs
Your staff are not doing anything strange. They want the reply written faster, and the AI helps. But the customer's name, email address and order details now sit in a personal account you do not control, under terms you never agreed to. This is often called shadow AI.
In Ireland and across the EU, you are responsible for what happens to your customers' personal data, including where a colleague pastes it. In the Emirates, the same question applies to the data you hold. I covered what to ask a provider in Where does your AI send your customers' data?
Banning AI rarely works. People switch to their phones, and you lose sight of it completely.
Three things that work better
- Give people a work account. A business plan from the AI provider, signed in with the work email, so the business terms apply and you can see and manage it.
- Turn off the settings that share chats. Check the activity and training settings on every work account, and write down what you chose.
- Write one rule everyone can remember. For example: no customer names, emails, phone numbers or invoices go into a personal AI account. Ever.
Google's page is the clearest I have read, but every personal AI app has its own version. Read the one your team actually uses. For the most sensitive work, see The most private AI never leaves the phone.
Do this today
- Ask your team which AI apps they used for work this month, and which account they were signed in to.
- Open the privacy page of the most used one and find how long it keeps chats and who can read them.
- Send everyone the one-line rule above.
Your next step
I write one practical note like this every day. Sign up just below to get them by email.
Drafted with AI help, edited and approved by me.