Questions to Ask Google Gemini About Yourself: 30 Prompts

ChatGPT knows you through what you type into it. Gemini, if you let it, knows you through what Google already has: the receipts in your inbox, the things you searched at 1am, the videos you fell asleep to, the photos you took and forgot about.
That is a very different starting point for the "ask the AI about yourself" game. A ChatGPT roast is built from your chats. A Gemini roast can be built from your life admin. That is why the scary questions here are actually scary, and also why most of the viral lists, which are ChatGPT lists with the name swapped, miss the point entirely.
So this guide is written for Gemini specifically. It covers what Gemini can and cannot see (and who actually has the feature, because a lot of readers do not), a quick test to tell a real read from a horoscope, and 30 questions sorted from fun to uncomfortable. Most of them only make sense on Gemini.
If you use ChatGPT more, we wrote the companion piece: questions to ask ChatGPT about yourself. The two pair well, and comparing their answers is one of the more interesting things you can do with an evening.
What Gemini can actually see about you
Gemini has three separate layers of personal context. Which ones are switched on decides which questions are worth asking.
| Layer | What it contains | Default | Where to find it |
|---|---|---|---|
| Past chats | Your earlier Gemini conversations | On | Settings, Personal context, "Your past chats with Gemini" |
| Instructions for Gemini | Things you asked it to remember ("remember I'm vegetarian") | Empty until you add some | Settings, Personal context, under the past chats toggle |
| Personal Intelligence | Connected Google apps: Gmail, Photos, Search, YouTube | Off, opt in | Settings, Personal Intelligence, Connected apps |
The first two are the same idea as ChatGPT memory. The third is the one that changes everything. Google launched Personal Intelligence in January 2026, and you choose which apps to connect. You can connect Gmail and leave Photos out, for example. Google's own launch example shows what that means in practice: asked about new tyres, Gemini pulled family road trips from Photos, read the licence plate off a picture, and found the car's trim level by searching Gmail.
Who actually has Personal Intelligence
This is the part most guides skip, and it matters if you are reading from London or Berlin. According to Google's overview page, at the time of writing (October 2026):
- It is rolling out globally except the European Economic Area, Switzerland, the United Kingdom and Nigeria.
- You must be 18 or over and use a personal Google account. Work, school and education accounts are excluded.
- In the US it opened to free users in March 2026. Elsewhere it reached Google AI Plus, Pro and Ultra subscribers in April 2026, with free access promised "in the coming weeks" (Android Authority).
- Google notes the Photos part is available to eligible users in the US, so outside the US you may get Gmail, Search and YouTube without Photos.
If you are in the UK or EU, skip down to the section for accounts without Personal Intelligence near the end. The fun and deep questions still work with past chats. The inbox ones will not.
The 60 second test: is it about you, or about everyone?
There is a well known psychology finding called the Barnum effect: vague statements like "you are hard on yourself but rarely show it" feel uncannily personal to almost everyone. We covered the original 1948 experiment in our ChatGPT version. The short version: a chatbot can produce lines like that all day, and they will land every time.
Gemini makes the check easier than ChatGPT does, because Google built two escape hatches into it:
- Regenerate without personalization. Google says you can regenerate a response without personalization. Ask your question normally, then regenerate it that way and compare.
- Temporary chat. Google says temporary chats "won't be used to personalize your Gemini experience." Ask the same question there.
If the personalised and unpersonalised answers say roughly the same thing, the first one was never about you. If the personalised one names a real email thread, a real trip, a real obsession you have been searching about, you are getting a genuine read.
Then add this to anything that matters:
For every claim you make about me, tell me where it came from:
a past chat, an email, a photo, a search or a YouTube video.
If it is a general guess, label it "guess."
Google says Gemini "will try to reference or explain" the personal information it used, and that you can ask for more if it does not. This prompt makes that the rule instead of a hope.
Google's own warning: your inbox is not your personality
Here is something you will not find in the viral lists, because it comes from Google itself. In the launch post, Google lists the ways Personal Intelligence gets people wrong:
- Over-personalization, where it links things that have nothing to do with each other.
- Missing timing and nuance, such as relationship changes. Google's example is divorce.
- Confusing presence with preference. Google's example: lots of photos at a golf course does not mean you love golf. You might just be driving your kid to lessons.
Think about what that means for the questions below. Your inbox is full of newsletters you never open, receipts for gifts, work threads that have nothing to do with who you are. Your search history includes everything you looked up for someone else. Gemini sees all of it with the same weight unless you tell it otherwise.
So before the scary questions, run this one:
List the five things you would assume about me from my connected
apps. For each, tell me which app it comes from. I will tell you
which ones are wrong.
Correct the wrong ones in the chat. Google says corrections like "I prefer window seats" carry into future answers, so a few minutes here makes every later answer better.

Thousands of photos say where you were. They do not always say why.
Fun questions to ask Gemini about yourself
Start here. These are low stakes and they show you quickly how much Gemini can actually see. If the answers reference real videos, real searches and real trips, Personal Intelligence is working. If they could describe anyone, it probably is not connected.
Roast me using my YouTube history and my searches. Keep it
affectionate, but make every joke specific enough that a stranger
could not have written it.
Based on my Google searches, what are my three most unhinged
interests? Tell me when each one started.
If my photos from the last year were a movie, what would it be
called and what genre would it be? Pick the five scenes that
would make the trailer.
Based on my emails and searches, what job would you guess I have,
and what job does my YouTube history suggest I secretly want?
Plan my perfect weekend using only what you can see in my
connected apps. Then tell me which detail you were least sure about.
What rabbit hole have I gone down most often this year, judging by
my searches and the videos I watch?
Based on my order confirmation emails, describe my shopping
personality in three words, then defend each word with a receipt.
What would my "year in review" look like if Google made one just
for me?
The order confirmations one is a good Gemini-only check. ChatGPT cannot answer it unless you have pasted receipts into a chat. Gemini, with Gmail connected, can.
"What do you know about me" questions
This is the audit. The point is to see what Gemini knows and, just as important, which app each piece comes from. Then compare it with your settings.
What do you know about me? Group it by source: our past chats,
my saved instructions, Gmail, Photos, Search and YouTube.
Keep it factual and say which source each item comes from.
What is something you know about me from my connected apps that
I never told you directly in a chat?
Which of your assumptions about me are you least confident in?
List them so I can correct them.
Which topics have I searched for or emailed about repeatedly over
the past year without ever seeming to finish with them?
If you had to brief a new assistant about me in five bullet
points, what would they be, and where did each one come from?
What could you NOT know about me, even with my apps connected?
The last one is a good reality check. Gemini's answer will usually be more honest about its limits than a list of what it does know, and the gaps (your offline life, conversations that never touched Gmail, anything you only said out loud) are a useful reminder of how partial the picture is.
Scary questions to ask Gemini about yourself
These are the ones that can genuinely sting. On ChatGPT, "what am I avoiding" is a guess built from what you chose to share. On Gemini with Gmail connected, it can look at the emails you actually never answered. That is a lot more uncomfortable, and a lot more useful.
Based on my Gmail, which emails or people have I been avoiding
replying to? Do not quote private details, just tell me who and
roughly how long.
Look at my order confirmations and subscription emails. What am I
paying for that I probably forgot about?
Compare what I search for with what I email about. Where do my
private curiosity and my public life not match?
What do my late night searches say about what is actually on my
mind? Point to the pattern, not to individual searches.
Based on the invites and booking confirmations in my Gmail and my
photos, how do I really spend my time compared to how I say I want to?
If someone who disliked me read my inbox and search history, what
would they say about me? Then tell me which parts are fair.
Based on my patterns across these apps, what is most likely to go
wrong for me in the next year if nothing changes?
What is one thing you noticed about me that you were not going to
mention unless I asked? I am asking.
Two cautions. First, Google says Gemini avoids making proactive assumptions about sensitive areas such as health, but will discuss them if you ask. So these questions can surface things it would normally leave alone. Second, a model asked to be brutal will try to deliver brutal. Treat every harsh line as a hypothesis to check against the source, not a verdict.
Deep questions to ask Gemini about yourself
These sit between fun and scary. They work best when you push back and answer the follow ups instead of just reading the reply.
Based on what I spend time and money on, what do I actually value,
as opposed to what I say I value in our chats?
Looking at my photos and emails, when did I seem happiest this
year? What was going on?
What strengths do I seem to underuse? Point to where you have
seen them, in chats or anywhere else.
What would the version of me from five years ago be most surprised
by, judging by how I live now?
The money question is where Gemini can genuinely beat a journaling app. Receipts are honest in a way that self-description is not.

Google's own launch example was a family trip planned from Gmail and Photos. The same data can plan a trip or describe a person.
No Personal Intelligence? (UK, EU, or a new account)
If you are in the EEA, Switzerland or the UK, or your account just has nothing in it yet, Gemini only has your chats to go on. Note that past chats personalization also reached free users later in Europe than elsewhere (9to5Google), so check Settings, Personal context to see whether the toggle is there for you.
You have two good options.
Option 1: let Gemini interview you. You supply the evidence, so the read is about you and not about a demographic.
Interview me. Ask me 10 questions, one at a time, that would tell
you the most about who I am. Do not summarise until all 10 are
answered.
Ask me about three decisions I made this year. Then tell me what
they have in common that I might not have noticed.
Option 2: bring your ChatGPT memory over. Since March 2026 Gemini has had an import tool in Settings: it gives you a prompt to run in ChatGPT or Claude, you paste the summary back, and Gemini saves it. Third-party guides report the same region limits as Personal Intelligence, so it may not be offered in the UK or EU either. If it is, it is a fast way to give Gemini years of context.
Then you can try the experiment we find most interesting:
Here is what ChatGPT says it knows about me: [paste it].
Based on our chats, what would you add, and what do you think
it got wrong?
Two models disagreeing about you is far more informative than either one agreeing with itself.
Before the scary ones: a quick privacy check
Asking Gemini to summarise your life pulls a lot of personal detail into one chat. Here is what is worth knowing, from Google's documentation and the Which? guide to Gemini's data settings:
- Your chats are kept for 18 months by default. Gemini activity auto-deletes after 18 months unless you change it to 3 or 36 months under Keep Activity (formerly Gemini Apps Activity). On a computer it lives at myactivity.google.com/product/gemini.
- Some chats are read by humans. A sample of chats is reviewed by human reviewers to improve quality. Google says reviewed chats are kept for up to three years and are disconnected from your account first. Deleting your activity does not delete those reviewed copies.
- "Off" still means 72 hours. With Keep Activity off, and in temporary chats, Google still keeps conversations for up to 72 hours to respond and keep the service safe.
- Gmail and Photos are not training data, the chat is. Google says Gemini "doesn't train directly on your Gmail inbox or Google Photos library." The prompts and responses themselves can be used in a limited, filtered form. So a reply that quotes your inbox is a different thing from the inbox.
- Screenshots travel. A "brutal truths" answer built from your Gmail can easily contain your employer, your bank or your ex. Read it before you post it.
- Disconnecting is one tap. Settings, Personal Intelligence lets you remove any app at any time, and Personal Intelligence is off until you switch it on.
If you only want to play with the fun questions, connect YouTube and Search and leave Gmail out. You lose the scary questions and keep most of the laughs.
Frequently asked questions
Can Gemini read my Gmail?
Only if you connect it. Personal Intelligence is off by default, and you choose which apps to link under Settings, Personal Intelligence, Connected apps. With Gmail connected, Gemini can search and reason over your emails to answer a question. Google says it does not train its models directly on your Gmail inbox.
How do I see what Gemini knows about me?
Check Settings, Personal context for your past chats toggle and saved instructions, and Settings, Personal Intelligence for connected apps. Then ask Gemini to list what it knows grouped by source. Comparing the two is the most complete picture you can get from inside the app.
Why are Gemini's answers about me so generic?
Usually because it has little to work with. Personal Intelligence may not be available in your country or on your plan, past chats may be off, or you may be in a temporary chat. In the UK and EU, the interview-style questions or importing your ChatGPT memory will get you further.
Is Gemini or ChatGPT better at knowing you?
It depends on where your life is. ChatGPT knows you through what you tell it, plus any apps you connect. Gemini with Personal Intelligence starts from your Google data, so it can know things you never mentioned. If most of your history is in ChatGPT, Gemini's memory import can narrow the gap.
Are temporary chats in Gemini private?
More private, not invisible. Google says temporary chats do not appear in your recent chats or activity, are not used to personalize Gemini or train its models, and are kept for up to 72 hours so Gemini can respond and process feedback.
How do I stop Gemini using my personal data?
Disconnect apps under Settings, Personal Intelligence, switch off "Your past chats with Gemini" under Personal context, and turn off or shorten Keep Activity at myactivity.google.com/product/gemini. Chats already selected for human review can be kept for up to three years, so turning things off only protects future chats.
The honest bottom line: Gemini with your apps connected is the closest thing yet to an AI that knows you without being told. That is exactly why its reads need checking. It sees everything you did, and very little of why you did it. Ask for the source, correct what it gets wrong, and the scary questions become useful rather than just unsettling.
We spend our days building AI systems that answer from real data instead of confident guesses. If you want that kind of rigour working inside your business, book a free call and we will show you where it pays off.
Image credits, all released under CC0 1.0 and hosted on Wikimedia Commons: "Black smartphone in hand" by Dennis Cortés, "Smartphone photographer" by James Sutton, and "Planning a trip over coffee" by rawpixel.com.