Google Gemini Now Lets You Import Chats and Memories from ChatGPT and Claude

Google just made switching AI chatbots feel like moving between streaming services rather than moving across the country. With a ZIP file and a quick copy-paste, your entire AI relationship history memories, preferences, thousands of conversations can now live inside Gemini. It’s the most aggressive user-acquisition play the chatbot wars have seen yet.

The AI chatbot market has a retention problem specifically, every platform has it. You spend weeks or months training your preferred assistant on who you are: your name, your job, your weird dietary preferences, the ongoing project you keep coming back to. Then you hear about a rival platform doing something interesting, and the thought of starting from scratch kills any curiosity before it can breathe.

Google knows this. On Thursday, the company announced what it’s calling “switching tools” for Gemini a pair of features designed to make the friction of changing AI assistants as painless as possible. You can now import your personal memories and preferences from competitors, and upload your entire chat history as a ZIP file. The message from Google is clear: switching costs are no longer an excuse to stay where you are.

The Problem Google Is Actually Solving

Here’s something the industry doesn’t talk about enough: AI assistants are becoming personalized in a way that makes them genuinely hard to leave. When you’ve spent months building context with a chatbot teaching it your communication style, your ongoing projects, your family members’ names that accumulated context has real value. It’s not lock-in in the traditional sense (no contracts, no payments lost), but it’s a psychological switching cost that’s just as powerful.

Google’s new tools are a direct answer to that problem. The memory import feature works through a clever, low-tech mechanism: Gemini generates a suggested prompt for you, you paste it into your current AI app (whether that’s ChatGPT, Claude, or any other major chatbot), and the rival app produces a summary of what it knows about you. You then copy that summary back into Gemini, which analyzes and stores it as structured memory.

It’s elegant because it doesn’t require any API access or cooperation from competitors. Google is essentially using the chatbots’ own intelligence against themselves asking them to summarize your preferences so Gemini can absorb them. No backdoors required.

How the Chat History Import Actually Works

The second feature is arguably more powerful than the memory import. Google is now accepting ZIP file uploads of full chat histories from other AI providers. If you’ve been using ChatGPT or Claude for a year and have hundreds of conversations stored, you can export that entire archive and bring it with you to Gemini.

Both ChatGPT and Claude support data export in ZIP format, so this isn’t a hypothetical it works today. Once uploaded to Gemini, your old conversations become searchable and usable as context. The company says users can “seamlessly pick up right where you left off.”

Here’s how the full switching process looks from end to end:

• Open Gemini Settings and look for the new “Import” option under memory or chat history.
• For memory: Copy the suggested prompt Gemini provides, paste it into your current AI app, and copy the generated summary back into Gemini.
• For chat history: Export your data as a ZIP file from your current chatbot, then upload it directly to Gemini.
• Gemini indexes everything your preferences become active memory, your old chats become searchable history.

What’s notable here is that Google is also renaming its existing “past chats” feature to simply “memory” as part of this rollout. That’s not just a cosmetic change it signals a shift in how Google wants users to think about Gemini’s relationship with their data. It’s not an archive. It’s a living, working memory that carries context forward.

The Numbers Behind the War for Chatbot Users

To understand why Google is making this move now, you need to look at the user numbers. The chatbot market is enormous and still accelerating, and the gap between ChatGPT and everyone else is real.

ChatGPT’s 900 million weekly active users versus Gemini’s 750 million monthly note those are different time windows, which makes the comparison less direct but still illustrative. OpenAI is measuring weekly engagement; Google is measuring monthly reach. Even accounting for the measurement difference, ChatGPT appears to have stronger daily habitual use, which is exactly what Google needs to chip away at.

Google has enormous structural advantages Gemini is the default assistant on Android devices and deeply embedded in Chrome and Google Workspace. Yet despite that distribution, consumer mindshare has stubbornly remained with ChatGPT. This switching tool is an attempt to attack the problem from a different angle: instead of hoping users discover Gemini organically, Google is making it actively easy to defect from the competition.

What This Means for the Broader AI Assistant Market

The strategic implications here go beyond Google vs. OpenAI. What Google has done is establish a precedent: AI chat data should be portable. You built that conversation history. Your preferences are yours. A platform that makes it hard to leave is implicitly saying your data belongs to them.

This is a page borrowed from a very different industry telecom. In the US, mobile number portability became mandatory in 2003, and it fundamentally changed how carriers competed. Before portability, the switching cost (losing your phone number) was so high that providers could treat customer retention as a lock-in problem rather than a value problem. After portability, they had to compete on service quality.

AI memory portability could have a similar effect. If users can move their context freely, platforms can’t rely on accumulated personalization as a moat. They have to keep earning your loyalty with better responses, better integrations, better value not just banking on the sunk cost of everything you’ve already trained into the model.

This directly connects to the ongoing debate around which AI assistant is actually worth using in 2026 a question that just got more complicated, because the answer is no longer tied to where you started.

The “Personal Intelligence” Play

The switching tools don’t exist in isolation they connect to a broader Gemini feature called Personal Intelligence, which Google introduced earlier this year. Personal Intelligence allows Gemini to draw on relevant context from your Gmail, Google Photos, Search history, and past Gemini conversations when generating responses.

The vision Google is building toward is an AI assistant that doesn’t just answer questions it knows you. Not in a generic, surface-level way, but with the kind of contextual depth that makes responses genuinely personalized. The imported chat history from another chatbot slots directly into this framework. Your old ChatGPT conversation about which neighborhood in Barcelona you were considering for a trip? Gemini can now use that when helping you plan your next Barcelona itinerary.

That’s a genuinely powerful use case. And it’s a compelling argument for making the switch, beyond any abstract loyalty to a particular platform. The value of an AI assistant compounds with context and Google is now offering to absorb contexts built elsewhere.

This kind of agentic, context-aware functionality is also what’s driving interest in Gemini’s task automation capabilities, which have been expanding steadily over the past few months.

Will OpenAI and Anthropic Respond?

The honest answer is: they probably have to. If Gemini is positioned as the platform that welcomes your data from anywhere, any platform that doesn’t offer equivalent portability starts looking like it’s holding your data hostage.

OpenAI already lets you export your ChatGPT data that’s a prerequisite for Google’s import feature to work. But offering an import tool that pulls from Gemini or other competitors? That’s a different calculation entirely. Anthropic, which makes Claude, is in a similar position. Both companies will watch how users respond to Google’s offering before deciding whether to match it.

The irony is that Google’s import feature works partly because OpenAI and Anthropic built good data-export tools. They made it easy to leave their platforms (in compliance with data portability regulations), and Google has used that openness as an acquisition channel. It’s a reminder that data portability features can cut in unexpected directions.

For developers and power users thinking about which AI platform to build habits around, this changes the calculus. As we covered in our analysis of OpenAI’s own superapp ambitions, the race to become your primary AI interface is intensifying and now it’s a race where past investment in one platform doesn’t have to be wasted.

Limitations and What You Should Know Before Switching

Beyond the geographic and account-type restrictions, there are practical questions worth thinking through before you make the jump. The memory import is only as good as what your current AI app knows about you if you’ve never used memory features in ChatGPT, for example, the generated summary might be sparse. The quality of what Gemini receives depends entirely on the quality of what the other platform has stored.

Chat history imports raise their own questions around data handling. Google says the import is “secure,” but users should understand that uploading a ZIP of potentially years of conversations is a significant data sharing event. If you’ve discussed sensitive personal matters, business strategies, or anything you’d prefer to keep private, take a moment to review your export before uploading it wholesale. You can always be selective export the conversations you want, not everything.

There’s also the question of model quality. Importing your history doesn’t guarantee Gemini will respond the way you’re used to. The AI’s personality, its reasoning style, its tendencies those are baked into the model, not imported from your data. You’re bringing your context; you’re still getting Gemini’s intelligence.

What’s Next for the Chatbot Wars

Google’s switching tools are a smart, well-executed move that will likely accelerate Gemini’s user growth among the portion of the market that was curious but deterred by switching costs. They won’t be the last word in this competition.

The next frontier is probably real-time cross-platform sync not just a one-time import, but ongoing memory that follows you wherever you chat. That’s harder technically and raises thornier privacy questions, but it’s where the user experience logic points.

What Google has done today is establish a principle: your AI context belongs to you, and it should move with you. That’s good for users regardless of which platform they ultimately choose. The more these companies have to compete on actual intelligence and value rather than accumulated lock-in, the better the products will get for everyone.

The chatbot wars are no longer just about which model is smartest. They’re about which platform you trust with your digital memory and today, Google made a strong argument for why that should be Gemini. Whether it’s enough to move users who’ve built years of habits with ChatGPT remains to be seen. But the barrier just got significantly lower.