Google and Anthropic Launch Competing Import Tools as AI Chatbot Data Portability Race Heats Up
Google Gemini and Anthropic Claude both ship tools to import rival chatbot data, signaling that user lock-in is now a front in the AI platform war.
Overview
The race to become the dominant AI assistant is no longer just about model quality. In March 2026, both Google and Anthropic rolled out tools designed to let users import their data from rival chatbots, turning data portability into the latest battleground in the AI platform war. Google launched “switching tools” for Gemini on March 26, allowing users to transfer chat histories and personal preferences from ChatGPT and Claude, as reported by TechCrunch. Anthropic had moved first, shipping a memory import feature for Claude on March 2 that works with ChatGPT, Gemini, and Microsoft Copilot, according to SiliconANGLE.
OpenAI, whose ChatGPT remains the market leader, has not announced a comparable import tool.
What We Know
Google’s Approach: Memories and Full Chat Histories
Google’s Gemini import system offers two distinct mechanisms. The first is a memory import: users copy a Google-provided prompt into their existing AI assistant, which generates a summary of stored preferences, writing style, and personal context. That summary is then pasted into Gemini’s settings. The second is a full chat history import, where users upload a ZIP file exported from ChatGPT or Claude directly into Gemini, making past conversations searchable within the interface, as reported by TechCrunch.
The chat history feature accepts files up to 5 GB in size, with a limit of five ZIP uploads per day. Processing times vary by provider and may take several days. The tools are free for all consumer Gemini accounts but are not available in the United Kingdom, Switzerland, or the European Economic Area, according to Bloomberg. Google has not disclosed a timeline for expansion to those regions. Only prompts and responses are carried over; project files, attachments, and AI-generated images cannot be imported.
Anthropic’s Approach: Memory-Only, on the Free Tier
Anthropic’s import tool, launched on March 2, takes a lighter approach. Rather than ingesting full conversation logs, it transfers only the distilled preferences and context that a competing chatbot has stored about the user. The process involves copying a prompt into a rival chatbot, which then generates a structured export of stored memories covering custom instructions, personal details, project goals, technical preferences, and behavioral corrections. That export is pasted into Claude’s memory settings, with the system taking up to 24 hours to fully assimilate the data, as reported by SiliconANGLE.
Notably, Anthropic made the memory feature available on Claude’s free plan alongside the import tool launch, according to 9to5Mac. The timing was not incidental: Claude had recently risen to the top of the iOS App Store rankings, and Anthropic positioned the import feature as a way to reduce friction for users migrating from competitors.
What We Don’t Know
Several important questions remain unanswered. It is unclear whether Google retains or uses imported chat histories for model training purposes, a distinction that carries significant privacy implications given the difference between importing raw conversation logs versus distilled preference summaries. The EEA exclusion from Gemini’s tools suggests potential regulatory concerns, but Google has not specified whether GDPR compliance is the obstacle or when the restriction might be lifted.
OpenAI’s silence on a comparable import tool is also notable. As the incumbent with the largest user base, ChatGPT has less incentive to facilitate inbound switching, but the absence of an export-friendly posture could become a liability if data portability expectations harden among users or regulators.
Analysis
The near-simultaneous launch of import tools by Google and Anthropic reflects a broader shift in how AI companies compete. When model capabilities converge, as recent benchmarks suggest they have across the leading frontier models, platform stickiness depends on accumulated context: the preferences, habits, and personal information that users have built up over months of interaction. By lowering the cost of switching, both companies are making a calculated bet that their products can win on quality even when users arrive with pre-existing context from a rival.
The two approaches also reveal different competitive philosophies. Google’s full chat history import brings raw data into its ecosystem, offering richer context but also raising questions about data handling. Anthropic’s memory-only approach is lighter and more privacy-conscious, importing only what the user’s previous assistant learned rather than every word exchanged.
For users, this is a net positive. The AI assistant market has been maturing rapidly, and the emergence of portability tools means that choosing an assistant is becoming less of a one-way commitment and more of a reversible decision, a dynamic that should keep all three major providers competing aggressively on both quality and trust.