Anthropic has announced that Claude now has a single memory available in the chat and Claude Cowork by merging those that both had separated until now.
The Artificial Intelligence (AI) laboratory has reported that Claude will allow more flexibility when using Claude Coworksince a task can be assigned based on the knowledge that the ‘chatbot’ has saved from previous conversations.
This means that the user will not need to repeat certain information, since Claude retains the context from the memory that is now unified, as he has shared on X (formerly Twitter).
This novelty comes after the launch of the Claude for Chrome extension, which has allowed Cowork to be activated in the Google browser, so now, all conversations on the side of the browser They will become part of that unified memory.
Likewise, Anthropic is giving give users more control over memory that Claude keeps. Information is organized by topic, which can be reviewed, edited or deleted, now included the option to include sensitive topics in memory.
In fact, this option can be activated from the welcome screen with Claude’s memory news when the app is opened, or from the memory settings in settings.
“By default, Claude does not save topics related to personal or sensitive topics such as health, race, ethnicity, religious beliefs, politics or gender identity and other similar areas,” explains Anthropic in the announcement published on its blog.
The AI laboratory points out that A notification will be sent to the user every time Claude saves a sensitive topicand that, to protect security and privacy, there is some data that will never be saved in memory, such as identification numbers, criminal records, immigration status, or anything that violates Anthropic’s Acceptable Use Policy (AUP).
Memory is activated by default in free, Pro and Max accounts in the web, desktop and mobile ‘apps’ versions. In the Team and Enterprise versions, it is the administrators who control their availability.