How to screen-share on Zoom & Google Meet without leaking secrets

You hit "Share screen" to show a demo or walk a colleague through something — and along with your slides, the call now sees whatever else is on your screen: a notification that just popped, the password manager you left open, a browser autofilling a token, an .env file in your editor. On a recorded meeting, that exposure is permanent, sitting in the cloud for anyone with the link.

Here's how to share safely on Zoom, Google Meet, and Teams — the rules to follow, where they fall short, and how to cover the gaps you can't watch yourself.

What screen-sharing actually exposes

• Notifications — Slack DMs, iMessages, email previews, and 2FA codes popping mid-share.
• Other windows & tabs you alt-tab to: a billing dashboard, an internal admin panel, a customer's data.
• Browser autofill — usernames, emails, and saved tokens that appear as you type.
• Your terminal and editor — API keys, .env values, connection strings.
• Password managers and email flashed for a second while you find something.

The rules to follow (every time)

1. Share a single window, not your whole screen. In Zoom/Meet/Teams, choose the specific app or browser tab — not "Entire Screen" — so nothing else can wander into view.
2. Turn on Do Not Disturb / Focus before you share, so notifications don't surface.
3. Close what you don't need. Quit the password manager, the billing tab, the email client before the call.
4. Use a clean profile. A separate browser profile (or macOS user) with no saved logins, history, or autofill for demos.
5. Pause sharing when you switch. Most apps let you pause the share while you move to something private, then resume.
6. Glance before you reveal. Check what's on screen before you bring a window forward.

Why the rules aren't enough

They all rely on you remembering, every single time, under the pressure of a live call. But "share one window" doesn't help when the secret is inside that window — your editor still shows the .env, your terminal still prints the key. A notification can fire in the half-second before Focus catches it. And on a recorded call, one slip is there forever. For sensitive data, you want a layer that catches what you miss — automatically.

Automate the safety net

That's why I built Censr. It watches your screen on-device (nothing is uploaded), detects API keys, tokens, passwords, and PII the moment they appear, and blacks them out before the frame reaches the call. Because it works as a virtual camera, you pick "Censr" as your camera/source in Zoom, Google Meet, Teams, or Discord — and it covers your whole screen, not just one app. It runs a short buffer so a freshly-revealed secret is sealed before it airs, plus a panic hotkey for everything else.

Honest about the one thing that matters: no detector catches 100% of everything. Censr is defense in depth — automatic detection + a safety buffer + a panic key + an always-blur list — layered on top of the habits above, not a replacement for them.