Just as OpenAI was facing a massive wave of cancellations due to its contract with the U.S. Department of Defense (#QuitGPT), Anthropic delivered the finishing blow. Anthropic had earlier announced a new version of its AI assistant, Claude."Memory Import" toolThis new feature allows users to seamlessly transfer personal preferences and conversational context accumulated on competing platforms such as ChatGPT, Gemini, or Copilot to Claude using a specific prompt.
Breaking free from "ecosystem hijacking": Three steps to painlessly transfer memories
In the past, the biggest pain point for users who wanted to switch their main AI assistant was that they had to rebuild their conversational rapport "from scratch".
Anthropic's new "Memory Import" feature allows users to extract all their commands, project details, and context without needing to understand code or wait for other platforms to open their API export interfaces. Simply copy a "memory extraction prompt" provided by Anthropic and paste it into ChatGPT or Gemini to force the old AI system to extract all your command preferences, project details, and context.
Finally, simply copy and paste the output into the "Manage memory" section in Claude settings.
Anthropic stated that Claude needs approximately 24 hours to process and integrate the newly imported context. Once complete, users can click the "See what Claude knows about you" button to confirm, and can also manually fine-tune the memorized content.
However, Anthropic also emphasized that Claude's memory function is mainly focused on "work-related topics to improve collaboration efficiency," so he may not deliberately retain overly private details that are not related to work.
Precise timing for the "home invasion": The #QuitGPT movement and the Department of Defense contract dispute
The timing of this new feature's release is definitely not a coincidence.
Recently, Anthropic's negotiations with the U.S. Department of Defense (DoD) broke down because it refused to remove its AI security barriers that "prohibit large-scale domestic surveillance" and "prohibit fully autonomous weapon systems." Subsequently,OpenAIThey accepted the classified military contract that Anthropic had rejected, which sparked a huge backlash on social media.
Dissatisfied with OpenAI becoming a "military-industrial complex," Western users launched a massive "#QuitGPT" boycott, with hundreds of thousands of users unsubscribing from ChatGPT. Anthropic seized this opportunity to release its "memory transfer" tool, practically shouting to these dissatisfied users who couldn't bear to part with their past AI memories: "The door is open, bring your luggage!"
With near-zero switching costs, Claude tops the App Store charts.
This combination of tactics has proven remarkably effective. Benefiting from the boycott and the seamless transition experience, Claude recently surpassed ChatGPT, which had long dominated the App Store's free app charts, to claim the number one spot.
Analysis of viewpoints
In the software services market, especially in the field of large language models, "switching cost" is the biggest moat for enterprises. You spend months or even years on ChatGPT to let it know who you are, what projects you are doing, and what kind of code layout style you prefer; once you switch platforms, all this "tacit understanding" will instantly disappear.
OpenAI has historically refrained from providing full memory data export functionality precisely to retain its heavy users. Anthropic's approach of extracting ChatGPT's memories through "natural language prompts" cleverly circumvents this technological barrier. This not only completely dismantles its competitor's competitive advantage but also successfully positions itself as a morally upright "anti-establishment hero" during the "#QuitGPT" public relations crisis.
In the future, the competition in generative AI will no longer be just a contest of model computing power; the battle for "digital memory sovereignty" is just beginning to intensify.
