Why Google Gemini Needs Your Search History (and Why I Won’t Share Mine)
On Thursday, Google made a number of previously paid Gemini features available to free users . You can now use Gemini’s own chatbots, which the company calls “Gems”; Deep Research, which runs artificial intelligence models that “think through” every step of a problem; and upload files to Gemini 2.0 Flash Thinking whether you pay for Google’s AI services or not.
But that’s not all: the company also introduced a new experimental feature for Gemini – Gemini with Personalization . Powered by Google’s Gemini 2.0 Flash Thinking model, this feature connects Gemini to your Google apps and services to offer you a more personalized AI assistant. The idea is that by connecting your Google account information to Gemini, it will learn more about you and be able to provide more informed results based on your personal tastes.
This is certainly a step in the direction that big tech companies are touting AI. But to work, you need to connect your search history to Gemini . That’s a lot of trust in Google’s AI service, and I think it’s a tough decision for anyone concerned about the amount of data we’re feeding these AI tools.
What can the Gemini personalization model do?
Google offers several examples of how this new service can improve your Gemini experience. You can ask the bot where you should go on vacation, and instead of pulling information from a number of sources about where other people like to vacation, the bot could theoretically use your past searches to zero in on a trip it thinks you’d like. Perhaps you’ve put together a vision board of sorts about a trip to the Bahamas or St. Lucia, and the bot will collect searches related to tropical vacations. Or maybe you’ll ask the bot to suggest a new hobby for you and see results based on the types of things you’ve searched for in the past.
I understand the concept that Google is going for: instead of having a bot that answers queries the same way for everyone, why not have each user’s bot provide answers tailored to their likes and dislikes? However, this makes me wonder: if a user is already looking for things like vacation spots and new hobbies, wouldn’t they be able to choose for themselves where they’d like to go or what kind of activity they’d like to do? If I’m doing a lot of research about running and ask the bot what hobby I should take up, I won’t be surprised when Gemini returns results for “On Shoes” and a local running club.
This isn’t necessarily a clever tactic on Google’s part. To use this feature, you need to connect your search history to Gemini. To me this is actually surprising and slightly refreshing. At least Google isn’t doing a default here.
Because the model is a “thinking” model, you will see the entire train of thought as part of the results. So Google says you’ll be able to see the personal information Gemini used to generate the response, including saved information, past conversations or search history. Additionally, Gemini won’t view your search history unless you specifically use this experimental personalization feature. (You’ll also need to turn on Google’s “Web and App Activity” setting.) All of this means that using this feature doesn’t seem like Gemini will crawl your search history every time you use it. If you’re using the standard Flash Gemini 2.0 model, it won’t pull this personal data along with the response—unless you switch back to Personalization.
Should you connect your search history to Gemini?
Here’s what I will say: I don’t connect my search history to Gemini—at least not yet. This feature is currently experimental, so this is not the full vision that Google has in store for it. (For example, the company plans to connect photos and YouTube data in the future.) But even if the feature is fully implemented, I just don’t feel comfortable connecting my personal search history to Google’s AI.
Don’t get me wrong: I know Google already has access to my search history (though disabling app and web search activity should reduce some data leakage). That’s not the point. Personally, I don’t feel the need to train Google’s AI to deal with my search history, which is what’s happening here. It’s a great idea to provide users with more personalized results from AI bots, but by enabling this feature, I’m providing free Google training for Gemini using my personal information. In fact, by requiring web and app activity to be enabled, Google is asking you to share this data with both Gemini and Google as a whole.
For all we know, Google may have the best intentions when it comes to privacy, but I nonetheless support a different AI decision with this decision: Don’t give personal information to AI . If you don’t want a Google reviewer to see what you share on Gemini, you probably shouldn’t share it at all. Traditionally, I’ve referred to this rule for things like confidential company information or highly personal information, but search history can also be highly sensitive. Do you really need a Gemini (or human reviewer) to see everything you were looking for just to try to make your Gemini results a little more personal? In any case, these results may be completely inaccurate .
How to use the Gemini personalization model
If you think these trade-offs are worth the potential benefits of Gemini’s personalization model, here’s how to try it. Open Gemini, then select “Personalization (Experimental)” from the drop-down list. Here Google will present you with a pop-up where you will need to connect your search history to Gemini. If you’re happy with this, select Connect Now .