# AI memory for teams

Teams can share [AI memory](https://www.knowman.io/guides/ai-agent-memory) today through version control and workspace settings: [Claude Code project instructions are shared with your team through version control](https://code.claude.com/docs/en/memory), and [ChatGPT workspace owners and admins can manage supported memory settings and role permissions](https://help.openai.com/en/articles/8590148-memory-in-chatgpt). [Anthropic's memory tool runs client-side, so your application controls where memory data is stored](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool). KnowMan is built for [one person's library](https://www.knowman.io/guides/personal-knowledge-base) today, and team plans are not available yet.

## What shared AI memory means

Shared [AI memory](https://www.knowman.io/guides/ai-memory-for-individuals) is information an assistant keeps between conversations and brings back when needed. Anthropic's memory tool lets Claude store and retrieve information across conversations in a directory of memory files ([Anthropic docs](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool)). In Claude Code, [memory and instruction files](https://www.knowman.io/guides/claude-code-memory) are loaded at the start of every conversation ([Claude Code docs](https://code.claude.com/docs/en/memory)).

Claude Code treats these files as context, not enforced configuration ([Claude Code docs](https://code.claude.com/docs/en/memory)). Anthropic also presents the memory tool as a way to build up a knowledge base over time ([Anthropic docs](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool)).

## Where team memory can live

With Anthropic's memory tool, your application holds the storage. The tool operates client-side: Claude requests file operations, your application runs them, and you control where and how the data is stored through your own infrastructure ([Anthropic docs](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool)).

Deleting a conversation is not the same as deleting memory. OpenAI notes that deleting the original chat does not automatically delete a separate saved memory ([OpenAI Help](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)). Deletion and Memory updates can also take time to propagate, and logs of deleted saved memories may be retained for up to 30 days ([OpenAI Help](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)).

## What Claude Code and ChatGPT share

Claude Code auto memory is machine-local. All worktrees and subdirectories within the same git repository share one auto memory directory, but files are not shared across machines or cloud environments ([Claude Code docs](https://code.claude.com/docs/en/memory)). [Project instructions](https://www.knowman.io/guides/teach-claude-code-your-knowledge) take a different path: they are shared with your team through version control, and the guidance says to focus on project-level standards rather than personal preferences in these files ([Claude Code docs](https://code.claude.com/docs/en/memory)).

ChatGPT Project-only memory draws a boundary too. Chats can reference other conversations in the same project, but not conversations outside it, and chats outside the project cannot reference conversations in it ([OpenAI Help](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)).

## Administering memory settings

ChatGPT workspaces place memory under owner and admin control: workspace owners and admins can manage supported memory settings and role permissions ([OpenAI Help](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)). OpenAI also notes that memory features and controls can vary by plan, region, platform, and workspace settings ([OpenAI Help](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)).

With Anthropic's memory tool, safety checks sit in your own implementation. Anthropic requires implementations to validate every path in every command to prevent directory traversal attacks ([Anthropic docs](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool)). A client-side implementation must also handle creating, reading, updating, deleting, and moving memory files ([Anthropic docs](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool)).

## Maintaining memory over time

Anthropic's memory tool supports creating, reading, updating, deleting, and moving memory files ([Anthropic docs](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool)). Anthropic recommends tracking memory file sizes, capping how large a file can grow, and periodically deleting files that have not been accessed in a long time ([Anthropic docs](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool)).

Retrieval happens on demand: rather than loading all relevant information up front, an agent records what it learns in memory files and reads them back on demand ([Anthropic docs](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool)). [ChatGPT memory](https://www.knowman.io/guides/ai-memory-for-individuals) does not retain every detail from every conversation, and what it uses can change as your context changes ([OpenAI Help](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)).

## Where KnowMan fits

KnowMan is built for one person's library today ([personal knowledge base guide](https://www.knowman.io/guides/personal-knowledge-base)). Team plans are not available yet.

For one reader's library, start with [persistent memory for AI agents](https://www.knowman.io/guides/persistent-memory-for-ai-agents) or [giving your AI context](https://www.knowman.io/guides/give-your-ai-context). To connect an assistant, see the [MCP server knowledge base guide](https://www.knowman.io/guides/mcp-server-knowledge-base).

## Questions

### Can a team share ChatGPT memory?

Within limits. ChatGPT memory can draw on saved memories, past chats, custom instructions, saved files, and connected apps, depending on plan and region ([OpenAI Help](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)). Workspace owners and admins manage [supported memory settings](https://www.knowman.io/guides/chatgpt-dots-memory) and role permissions, and Project-only memory keeps context inside its project ([OpenAI Help](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)).

### Is Claude Code memory shared across machines?

No. Auto memory is machine-local, shared only across worktrees and subdirectories of the same git repository ([Claude Code docs](https://code.claude.com/docs/en/memory)). Project instructions are the team-shared part, distributed through version control ([Claude Code docs](https://code.claude.com/docs/en/memory)).

### Who controls the data in Anthropic's memory tool?

Your application does. The tool runs client-side, so Claude requests file operations and your application executes them, with storage in your own infrastructure ([Anthropic docs](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool)). Anthropic says this can support building up a knowledge base over time ([Anthropic docs](https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool)).

### Does deleting a chat delete its saved memory?

Not automatically. OpenAI states that deleting the original chat does not automatically delete a separate saved memory ([OpenAI Help](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)). Deletion and Memory updates can take time to propagate, and logs of deleted saved memories may be retained for up to 30 days ([OpenAI Help](https://help.openai.com/en/articles/8590148-memory-in-chatgpt)).

## Related guides

- [What is a personal knowledge base, and how do you build one?](https://www.knowman.io/guides/personal-knowledge-base)

- [Persistent Memory for AI Agents](https://www.knowman.io/guides/persistent-memory-for-ai-agents)

- [What Is a Knowledge Base for AI Agents?](https://www.knowman.io/guides/knowledge-base-for-ai-agents)

- [Connect your knowledge base through KnowMan's MCP server](https://www.knowman.io/guides/mcp-server-knowledge-base)

- [How to Give Your AI Context from What You Read](https://www.knowman.io/guides/give-your-ai-context)

## More in this topic

- [How Can AI Remember What You Read?](https://www.knowman.io/guides/ai-that-remembers-what-you-read)
- [How to keep notes with a clear source trail](https://www.knowman.io/guides/chatgpt-note-taking)

## Start with your own library

KnowMan is built for one person's library today, and you can try Pro free for 14 days with no credit card. [Create your account](https://app.knowman.io/signup)

[Start free](https://app.knowman.io/signup)

Every new account starts with 14 days of Pro. No credit card.

Source: https://www.knowman.io/guides/ai-memory-for-teams