You can't stop screenshots — control what's on the screen instead

• Security
• Product

Every data room vendor eventually fields the same request: can you stop people screenshotting our documents? The honest answer is no — and any vendor who says otherwise is selling you theatre.

Why screenshot blocking is a losing game

In-browser "screenshot protection" works by detecting key combinations, blacking out the page on focus loss, or stripping the surrounding chrome. These are browser tricks, and browser tricks live inside the browser. The threat doesn't. A reviewer can point a phone camera at the monitor, use the operating system's own capture tool, or photograph the screen from across the desk. No JavaScript reaches any of those. Treating client-side blocking as a real control gives teams false confidence about a boundary that was never enforceable.

So we don't pretend. We spend the effort where it changes the outcome: controlling what is actually on the screen.

Burn the identity into the pixels

If you can't stop the capture, make the capture incriminating. Every page a viewer sees carries their identity — name, email, IP, timestamp — woven across the document itself, not floated on top as a layer the client could peel away. We do this server-side, on both the view path and the download path, so the watermark is part of the bytes the viewer receives. It survives a photograph because it is the photograph: a phone snap of a leaked page still shows exactly who was looking at it.

This matters most for the documents that leak: PDFs, and Office files, which we convert to PDF and then burn the watermark into the same way. The control isn't optional client chrome — it's baked into the rendition before it ever reaches the browser.

We're precise about the edges, because honesty is the point. Spreadsheets are deliberately handled differently: converting them to flat pages would destroy the formulas and structure that make a spreadsheet useful, so we apply a recipient-identifying forensic marker inside the workbook — a genuine deterrent and a forwarding trace, but one a determined user could strip. And our opt-in live editing mode, where the document renders natively rather than as a watermarked image, is governed by a cryptographically signed permissions set that the client cannot forge. We'd rather tell you exactly where attribution is strong and where it's a deterrent than overstate a guarantee.

Remove what shouldn't be seen at all

The strongest control is the one most often skipped: don't put sensitive content in front of the viewer in the first place. Redaction removes what a counterparty has no business seeing — names, figures, terms — before the page is ever served. A screenshot of redacted content reveals nothing, because there is nothing there to capture.

Prevention you can't enforce is theatre. Attribution that survives a photo, and removal of what shouldn't be shared, are controls you can stand behind. That's where we put the work.