Google、Geminiのメモリ機能とインポートツールを欧州に展開

Google wants Gemini to remember you and to lower the cost of switching

Google is expanding Gemini’s personalization features in Europe, bringing a memory system that can retain details from previous conversations and use them to tailor future responses. At the same time, the company is adding import tools designed to let users bring context from other AI assistants into Gemini by uploading chat-history files or pasting summary prompts.

Taken together, the two features represent more than a product update. They show how consumer AI competition is shifting from one-off chat quality toward continuity, retention, and switching friction. The question is no longer only which assistant answers best in the moment. It is increasingly which one becomes most useful over time without trapping users in isolated conversation histories.

What the new memory feature does

According to the report, Gemini’s “Memories” feature is rolling out to all users in Europe in the coming weeks after having previously been available to US users. The system is on by default, though users can turn it off in settings.

Memory allows Gemini to retain information such as a user’s name, job, hobbies, or location and to draw on those details when the model judges them relevant to a later conversation. That puts Gemini closer to the kind of persistent interaction model that has become increasingly important across mainstream AI products. Instead of treating every session as a blank slate, the assistant can build a working profile of user preferences and background.

In practical use, that can make responses feel more consistent and reduce repetitive prompting. A user who has already established preferred writing style, professional context, or recurring tasks does not need to restate those details every time. That improves convenience, but it also raises the strategic value of stored user context.

Why the import tools may matter even more

The more quietly consequential feature may be Gemini’s new import capability. Google is giving users two ways to move context from another AI app: upload a ZIP archive of chat history or paste an import prompt that summarizes relevant preferences and patterns from another assistant.

That is notable because it directly addresses one of the biggest practical barriers to switching between AI products. Users often invest time teaching a model how they work, what they prefer, and what projects they are juggling. Even if a competing assistant seems better, moving means losing much of that accumulated context. Import tools turn that lock-in problem into a product battlefield.

By making migration easier, Google is effectively arguing that personalization should be portable. That is useful for users, but it is also plainly competitive. Gemini is not just trying to remember existing users better. It is trying to make it simpler for someone else’s user to arrive with their prior context intact.

Consumer AI is moving from chat apps to relationship platforms

The broader shift here is that assistants are starting to behave less like search boxes and more like ongoing software relationships. Memory increases stickiness because the system improves as it learns a person’s recurring needs. Import tools reduce the cost of breaking that stickiness. Together, they define a new front in platform competition.

This has two implications. First, product differentiation now depends partly on how well a system manages long-term context, not just raw model performance. Second, control over user data and preference histories becomes more central to product strategy, trust, and regulation.

The European rollout is especially significant because personalization features often attract closer scrutiny in privacy-conscious markets. The report notes that memory can be disabled, which gives users some control. Still, as assistants store more persistent context, questions about visibility, consent, and lifecycle management will only grow more important.

Google’s timing reflects a more mature AI market

This update lands at a moment when the consumer AI field is becoming less about novelty and more about workflow capture. The major players are trying to become default companions for writing, planning, ideation, and everyday knowledge work. In that environment, remembering user preferences and importing historical context are not side features. They are part of the foundation of product adoption.

For Google, Europe is also an important proving ground. A successful rollout would show that richer personalization can expand internationally while still giving users settings-based control. For users, the promise is convenience and continuity. For rivals, the message is sharper: memory quality and portability are now competitive requirements.

Gemini’s new tools do not settle the broader debates around AI privacy or user lock-in. But they do make one thing clear. The next stage of AI competition is not just about who can answer your prompt. It is about who can remember enough to be useful tomorrow, and whether you can take that relationship with you if you decide to leave.

This article is based on reporting by The Decoder. Read the original article.

Google is bringing Gemini’s memory features to Europe and making AI switching easier

Google wants Gemini to remember you and to lower the cost of switching

What the new memory feature does

Keep Reading

OpenAIのAWS進出は、クラウドAIの力関係に大きな転換をもたらした

Why the import tools may matter even more

Consumer AI is moving from chat apps to relationship platforms

Google’s timing reflects a more mature AI market

NewsGuardの監査で、MistralのLe Chatがイラン戦争に関する誤情報プロンプトに脆弱であることが判明

Comments (0)