Claude Now Lets You Import AI Memory from Competitors in Under a Minute
Original: Switch to Claude without starting over View original →
Claude Introduces AI Memory Import
Anthropic has launched Claude Memory Import, a feature that allows users to transfer their accumulated preferences and context from other AI services — including ChatGPT — to Claude in under a minute.
How It Works
The process is deliberately simple:
- Copy a prompt: Paste a Claude-provided prompt into your existing AI chat to extract your preferences and context in a single conversation
- Paste into Claude: Drop the results into Claude's memory settings — Claude automatically updates its memory
The entire process takes under a minute and preserves months of accumulated context.
Availability
The feature is available across all Claude paid plans (Pro, Team, and Enterprise).
Strategic Significance
This feature directly lowers the switching cost from competing AI services. It arrives at a particularly significant moment: following OpenAI's deal with the U.S. Department of Defense, a wave of users are publicly canceling ChatGPT subscriptions and moving to Claude. The memory import feature removes a key friction point in that transition. The Hacker News post scored 396 points, reflecting strong community interest in the capability.
Related Articles
Why it matters: persistent memory is one of the missing pieces between demo agents and useful long-running agents. Anthropic pushed the feature into public beta on April 23 and framed it as a memory layer that learns from every session.
AI model availability is being shaped by export-control decisions, not only product readiness. Anthropic said it received notice that the U.S. Department of Commerce lifted export controls on Claude Fable 5 and Mythos 5, with access restoration starting the next day.
Anthropic says Claude contains a J-space that resembles a global workspace for active, verbalizable thoughts. The lead tweet has more than 9.1 million views and points to audit use cases, including hidden goals in sabotage-trained models.