Whether you're looking for an angel investor, a growth advisor, or just want to connect — I'm always open to great ideas.
Get in TouchGoogle rolls out dual import features in Gemini: chat history via .zip file (up to 5GB) and memory via summaries. Seamless migration from ChatGPT, Claude, and others. Unavailable in EU due to data regulations. The most aggressive AI platform consolidation play yet.
AI, startups & growth insights. No spam.
Google just made the most aggressive move yet to pull users away from ChatGPT and Claude. As of March 26, 2026, Gemini can import your full conversation history and personal memory from competing AI platforms — eliminating the single biggest reason people stay locked into a chatbot they've outgrown.
The feature is live now. It works. And it changes the competitive calculus for every AI company overnight.
In almost every mature software category, switching costs are the moat. It's why enterprise customers stay on Salesforce despite hating it. It's why users keep their Apple IDs even when Android looks tempting. And it's precisely why AI chatbot users — who've spent months teaching ChatGPT their preferences, their writing style, their workflows — don't switch platforms even when a better option exists.
Google just dismantled that moat.
The new Gemini import tools, reported by 9to5Google and covered by MacRumors, give users two distinct mechanisms to carry their AI history from ChatGPT, Claude, and other platforms directly into Gemini. Not a pale summary — your actual conversations, your actual context, your actual preferences.
This isn't a quality-of-life improvement. It's a market share play.
Google is essentially saying: the history you've built elsewhere doesn't have to die when you leave. Come to Gemini, bring everything. The barrier to switching just dropped to near zero — at least for users outside the EU.
The move comes as the AI assistant market is entering a critical consolidation phase. GPT-4o, Claude 3.7 Sonnet, and Gemini 2.0 are converging on quality. When raw capability gaps narrow, distribution and retention mechanics decide who wins. Google is betting that eliminating switching friction — combined with deep integration across Gmail, Docs, Search, and Android — is enough to break OpenAI's momentum and blunt Anthropic's enterprise push.
Google built two separate import pathways, each targeting a different layer of what makes a chatbot feel personalized.
This method transfers your actual conversation archive — the full back-and-forth threads you've had with competing AI platforms. Here's the exact flow:
Step 1 — Export from your current platform:
Step 2 — Open Gemini import:
gemini.google.com/import.Step 3 — Upload:
Once imported, your chats are fully searchable within Gemini. You can delete individual conversations or wipe an entire import batch. You can pick up threads, reference past decisions, and build on prior context — all without re-explaining your situation from scratch.
The second pathway is more clever — and arguably more powerful. Instead of raw chat logs, it transfers your contextual identity: who you are, what you care about, how you work.
The process: Gemini provides you with a structured prompt. You copy that prompt, paste it into ChatGPT or Claude, and the source AI generates a preferences summary across five defined categories:
You copy that summary back into Gemini. From that point forward, Gemini knows the same key facts your previous AI knew — without requiring weeks of re-training through natural conversation.
The import is substantive, but it has real limits. Understanding the gap matters before you migrate.
What transfers cleanly:
What doesn't transfer:
The memory import is also limited by the quality of the source AI's summary. If your existing chatbot has a shallow understanding of you, that shallowness is what Gemini inherits. Garbage in, garbage out — the import doesn't synthesize better context than what was already there.
Still, for the typical power user who has spent months building conversational history with ChatGPT or Claude, the import captures the most practically valuable layer: the accumulated knowledge of their work, preferences, and communication style.
There's a hard geographic wall on this feature — and it runs along European data law.
The Gemini import tools are unavailable in the European Economic Area (EEA), Switzerland, and the United Kingdom. Users in those regions cannot access the import interface, regardless of their Gemini subscription tier.
The reason is regulatory, not technical. The EEA is governed by GDPR — the General Data Protection Regulation — which imposes strict requirements on how personal data is processed, transferred, and repurposed. Uploading conversation histories from one AI platform into another raises non-trivial questions under GDPR: Is this a new processing purpose that requires fresh consent? Does Google's use of imported data comply with the original consent users gave to OpenAI or Anthropic?
The UK's post-Brexit data protection framework — still closely modeled on GDPR — raises similar issues. Switzerland's Federal Act on Data Protection (nFADP), updated in 2023, adds another layer.
Google's decision to simply exclude these regions rather than attempt compliance suggests one of two things: either the legal complexity is genuinely unresolvable in the near term, or the business priority for EU-based user acquisition doesn't yet justify the legal investment. Possibly both.
For European Gemini users, this creates a frustrating asymmetry. The competitive friction that Google is eliminating for American and Asian users remains fully intact in Europe. ChatGPT and Claude continue to benefit from data portability barriers that Google can't legally dismantle — at least not yet.
This exclusion also signals something broader about the regulatory divergence between AI markets. As the EU's AI Act implementation accelerates and GDPR enforcement tightens, features like this will increasingly bifurcate global AI products into EU-compliant and non-EU versions. European users will get safer — but less capable and less portable — AI experiences by default.
In technology markets, switching costs aren't just a business metric — they're a structural force that shapes entire industries.
Microsoft understood this when it bundled Office. Adobe understood it when it moved to Creative Cloud subscriptions. Apple understood it when it created iMessage. Lock users into a data ecosystem, make their history valuable, and exits become painful. That pain is retention.
OpenAI has benefited enormously from this dynamic. ChatGPT users who've spent six months building custom instructions, populating GPT memory, and developing a working relationship with the model's behavior don't leave easily. The cost isn't financial — it's epistemic. Starting over means re-teaching an AI everything it already knows about you.
Google's import tool attacks that lock-in directly. By making ChatGPT and Claude history portable into Gemini, Google converts the competitor's data asset into a Google onboarding tool. Every conversation you had with OpenAI becomes a reason to switch to Google, not a reason to stay.
This is strategically brilliant — and it sets a precedent. If Gemini imports succeed in driving meaningful user migration, expect OpenAI and Anthropic to respond with their own import tools targeting Gemini users. The resulting data portability arms race could paradoxically benefit users: if everyone can import from everyone, platform history stops being a lock-in mechanism entirely, and competition shifts back to raw capability and user experience.
Google has a unique structural advantage in this play. Gemini is already deeply embedded in Gmail, Google Docs, Google Search, and Android. Importing a ChatGPT history into Gemini isn't just changing chatbots — it's plugging into a productivity ecosystem that OpenAI cannot replicate. The chatbot is the entry point; the ecosystem is the lock-in. Read more about how Google is building this integrated AI layer in our analysis of Google Flow and Gemini's creative studio ambitions.
OpenAI and Anthropic are facing an uncomfortable reality: Google just made their user data a migration asset for a competitor.
For OpenAI, the immediate concern is GPT-4o's positioning. OpenAI has argued that ChatGPT's edge comes from its memory system, its custom GPT ecosystem, and its integrations. Google's import tool doesn't eliminate those advantages — custom GPTs don't transfer, and OpenAI's plugin ecosystem has no Gemini equivalent. But it does remove the inertia advantage. Users who were staying because leaving felt expensive now face a much lower barrier.
Anthropic is in a slightly different position. Claude has cultivated a strong following among professional writers, researchers, and developers who value its precision and its longer context windows. Those users are less likely to switch based on import convenience alone — their preference is often genuinely model-quality driven. But Anthropic's enterprise ambitions require sustained user growth, and any feature that makes Claude feel like a leaky container rather than a durable platform is a risk.
The deeper strategic problem for both companies: they cannot easily replicate Google's response. If OpenAI builds an import tool for ChatGPT that pulls from Gemini, it helps Google as much as OpenAI — it normalizes the idea that chatbot history should be portable, which reduces lock-in everywhere. The only winning move is to out-compete on capability and ecosystem, not on data retention.
Both companies will need to accelerate the parts of their platform that Google can't replicate through an import button: specialized models, unique integrations, and differentiated reasoning capabilities.
Google's move is being framed as pro-user. And in many ways, it is. Data portability — the ability to move your information between services — is a genuine consumer right that technology platforms have historically resisted.
But the framing deserves scrutiny.
Google is selectively pro-portability. When it benefits Google — importing competitor data into Gemini — data portability is a feature worth building and marketing. When portability would hurt Google — exporting Gmail data, Google Photos libraries, or Google Workspace histories to competitors — the company's enthusiasm has historically been more muted, despite Google's nominal participation in the Data Transfer Project.
This isn't unique to Google. Every technology company is selectively pro-portability in the direction that serves their growth. The question is whether regulators or competitive pressure will eventually force genuine bilateral portability across AI platforms.
For users, the practical question is simpler: does this feature make your life better? The answer is largely yes. Being able to carry your AI history without starting from zero is a genuine quality-of-life improvement. The fact that Google benefits from it doesn't make it less useful to the person doing the import.
The data also doesn't disappear from the source platform when you import it into Gemini. You're copying, not moving. Which means users who import their ChatGPT history into Gemini still have that history in ChatGPT. This isn't a zero-sum exit — it's additive, at least initially.
Whether Google uses imported conversation data to train its models is a separate, critical question. Google's privacy policy governs what happens to uploaded data — and users should read it carefully before uploading years of personal conversation history to a new platform.
If you're a heavy ChatGPT or Claude user considering this, here's a grounded assessment.
The chat import works as described. Uploading your .zip archive will surface your conversation history in Gemini's interface — searchable, browsable, usable as reference. If you've asked ChatGPT something complex and want to continue that thread in Gemini, you can find the original conversation and reference it directly. That's genuinely useful.
The memory import is more variable. Its quality depends on how much context your current AI actually has about you. If you've never set up ChatGPT's custom instructions or had substantive personal exchanges, the preference summary will be thin. If you've been detailed and consistent in your AI interactions, the summary will carry real value.
Don't expect seamless continuity. Imported chats are reference material — they don't automatically influence Gemini's behavior the way that Gemini's native memory system does. You're not teleporting a trained relationship; you're bringing a transcript. The relationship itself has to be rebuilt through actual use.
The 5 GB file size limit is generous for most users — OpenAI and Anthropic conversation exports rarely approach that size. The 5 .zip uploads per day allows for staged migrations if you're managing multiple export archives.
Plan your migration with Google's broader ecosystem in mind. If you're already using Gmail, Google Docs, or Android, Gemini's integration advantages become much more tangible post-import. If you're primarily on Apple devices with iCloud workflows, the ecosystem pull is weaker — though Google is actively closing that gap. See our breakdown of Apple's AI strategy and how Gemini fits into the iOS landscape.
Google's Gemini import feature isn't a minor update. It's a structural intervention in the AI market — one designed to collapse the switching costs that have kept users anchored to ChatGPT and Claude.
The mechanics are clean: export your archive, upload to Gemini, carry your history forward. The memory prompt method adds a second vector, transferring not just conversations but the personal context that makes a chatbot actually useful. Together, the two tools remove the most practical barrier to AI platform migration that has existed since ChatGPT's launch in late 2022.
The EU exclusion is a meaningful limitation — and a signal that data portability features will continue to bifurcate between regulated and unregulated markets. European users face a different, more constrained AI competitive landscape for the foreseeable future.
For OpenAI and Anthropic, the pressure is real. Their moats just got shallower. The response will have to come from capability, not inertia — from building AI that people actively want to use, not AI that users are trapped inside.
For Google, this is the right move at the right moment. Gemini's quality improvements over the past year have narrowed the capability gap with GPT-4o. Eliminating switching friction converts that parity into opportunity. If users can move without pain, some of them will — and Google's ecosystem depth gives those switchers a reason to stay.
The AI switching war has started. The first shot was a .zip upload.
Sources: 9to5Google — Gemini Import Feature | MacRumors — Gemini Import Tool | TechCrunch — Chatbot Switching Tools
Meta pushes back its next-gen Avocado model to May after underperforming rivals, and its AI leaders are now discussing temporarily licensing Gemini from Google — putting the $135 billion AI bet under scrutiny.
The ARC-AGI-3 benchmark replaces static puzzles with interactive environments requiring real-time learning. GPT-5.4, Claude Opus 4.6, and Gemini Pro all score below 1%. Humans score 100%. The implications for AGI research are massive.
Google Research unveils AI Co-Scientist, a multi-agent system built on Gemini 2.0 that autonomously generates, evaluates, and refines biomedical hypotheses — promising to compress years of research into weeks.