Choosing a data room: zero trust document access and monitoring
A data room is a closed online environment in which an organisation shares confidential documents with a limited group of outsiders: investors during a fundraise, buyers in an acquisition, auditors. Almost every provider promises security. The difference lies in three questions: who can see what, how do you know who saw what, and what happens if something gets out anyway.
Zero trust: nothing is shared until it is shared
Zero trust document access means nobody can see anything unless it has explicitly been shared with them, and the server checks that again on every request. Not once at login, but on every preview, every download and every search.
In practice you share with groups, not with single documents: a group per investor or per type of investor, and a folder shared with a group also covers everything added to it later. That works, but it does mean you design the folder structure around who may see what.
A detail that matters: what has not been shared should not exist for the reader. No folder names that are visible but closed, and a request for a document without access should say it does not exist rather than "access denied". Otherwise the data room itself gives away which documents are there.
The AI chat is a way in too
More and more data rooms have an AI chat that lets investors ask questions of the documents. That is an extra door to the same content. The chat may only answer from documents the asker may open, and that filter belongs in the search step, not afterwards. Suggested questions and search results can also reveal that a document exists. We described how to build that in our article on AI chat over confidential documents.
Monitoring: who saw what, and when
Monitoring document access starts with an audit trail that records every action: who, in which role, what, when. Think of previews, downloads, folders shared and unshared, publications, questions to the chat and login attempts.
A log on its own helps little if nobody looks at it. So also ask what the administrator can do with it: see per investor which documents were viewed, which questions come up most, and export the whole lot for a file or an audit.
If something gets out anyway
No data room stops someone photographing a screen. What does help is a dynamic watermark: the reader's name, email address, date and time on every page they open or download, plus a unique access ID that is also in the audit trail. If a page turns up somewhere, you can trace who viewed it and when. That deters, and it makes a leak traceable.
Accounts: MFA and sessions
The weakest point of a well-protected data room is a stolen password. Two-factor authentication should be mandatory for everyone who manages documents, and preferably for investors too. You also want users to see and end their active sessions, and to get a notice when a new device is added.
Questions for a data room provider
- Is access checked on the server on every request, or only at login?
- Can you share per group and per folder, and what happens to documents added later to a shared folder?
- Does a reader see folders or documents they have no access to, even if only the name?
- If there is an AI chat: does it only answer from what the asker may read, and how was that tested?
- What is in the audit trail, can you see per person what was viewed, and can you export it?
- What does the watermark look like, and is it on every preview and download, including zip files?
- Can you switch off downloads per document or per group, if you only want to give view access?
- Can MFA be enforced, for which roles, and can users manage their sessions?
- Where are the documents and the AI processing, and does data go to an external AI service?
On that last question, do not accept "secure" or "encrypted" as the answer. Ask which service does the AI processing, in which region, and what happens when it is unavailable.