Revoking Access When Someone Leaves The Team
The seat gets deleted. The access doesn’t.
Somebody leaves the team. You go into your anti-detect browser’s dashboard, remove them from the workspace, and move on. That’s the instinct, and it’s wrong in a way that matters if you’re running shared profiles across a team.
Removing a user from a workspace in Multilogin, GoLogin, Kameleo, AdsPower, or Dolphin Anty revokes their ability to log into the dashboard and pull new profiles down. It does not touch anything already on their machine. If they exported a profile, cloned one locally, or just left their laptop logged into an active session, none of that goes away when you click “remove member.” You’ve locked the front door and left three windows open.
This is worth being precise about because shared profile access control is not one control, it’s several, and they fail independently.
What a workspace removal actually does
These tools separate two things: the account layer (who can log into the dashboard, what permissions they have) and the profile data layer (the actual browser fingerprint config, cookies, local storage, and session tokens for each profile). Team management features in these platforms control the account layer. When you remove someone, you’re revoking their API key or session token for the management console. Their dashboard login stops working. Any profile they haven’t already synced or downloaded, they can’t get.
But the profile data itself, once it’s been opened on a machine, exists in two places: the cloud copy tied to the profile ID, and a local cache on whatever device ran it. Most of these tools work by syncing a profile’s state (cookies, local storage, sometimes extension data) up to the cloud when you close it, so the next person who opens that profile picks up where the last session left off. That sync is the whole point of “shared” profiles for a team. It’s also exactly why revoking dashboard access isn’t the same as revoking data access.
If someone had local admin or export permissions and pulled a profile down as a standalone folder or zip before you removed them, that local copy is a snapshot. It has whatever cookies and session tokens existed at export time, and it will keep working on the platform it’s logged into until those sessions expire or you force a logout server-side, independent of anything you do in the anti-detect tool’s dashboard.
Session tokens don’t check who’s holding them
This is the part people underestimate. A logged-in session on a platform like a marketplace, ad account, or social network is represented by a cookie or a bearer token, not by a check against your team roster. The platform doesn’t know or care that the person using that session got removed from your Kameleo workspace an hour ago. As far as the platform is concerned, it’s a valid session making requests.
That token stays valid until one of three things happens: it expires on its own (some platforms expire idle sessions after weeks, some barely expire at all), the platform detects something suspicious and force-logs-out the session, or someone with access to the actual account changes the password or explicitly revokes active sessions from the account’s security settings. Nothing in your anti-detect browser’s team management touches that third mechanism. You have to go into the underlying account and kill the session yourself.
If a former teammate still has a synced local profile with a live session cookie, they can keep using that account until you go log into it directly and force a password reset or session revoke. That’s not a flaw in the anti-detect tool. It’s just how session cookies work everywhere, and the anti-detect layer doesn’t and can’t override it.
Where credentials actually live
Beyond session cookies, there are usually several other places access persists:
Saved passwords in the profile’s browser storage. If your team uses the browser’s built-in password manager inside profiles (rather than a separate vault), those saved credentials travel with the profile data. A local export takes them along too.
Proxy credentials attached to the profile. Most of these tools let you bind a proxy (with its own username and password) to a profile config. If that proxy credential was visible to the person who had the profile open, or saved in a config file they exported, they still have working proxy access until you rotate the credential on the proxy provider’s side, not just remove them from the browser workspace.
API keys used for automation. If you’re driving profiles through Multilogin’s or GoLogin’s API for anything scripted, whoever had the API key or agent token has it in whatever script or config file they wrote it into, on whatever machine they wrote it on. Deleting their dashboard user doesn’t invalidate a key that was already generated and copied elsewhere unless you specifically go revoke that key.
Extension data and 2FA authenticators. If a profile has a password manager extension or a TOTP-generating extension logged in, that extension’s stored credentials are part of the profile’s local storage, same as cookies.
A practical offboarding sequence
None of this is about the anti-detect tool being poorly designed. It’s about treating “remove from workspace” as step one of an offboarding process, not the whole process. A reasonable sequence looks like:
- Revoke workspace access first, so no new profiles can be pulled and no new sync can happen.
- Rotate credentials on every underlying account the person had access to through those profiles, not the browser tool itself. Force logout of active sessions where the platform supports it, and change the password.
- Rotate proxy credentials separately from the browser workspace, on the proxy provider’s dashboard, since the proxy auth is independent of the anti-detect tool’s access controls.
- Revoke or regenerate API keys if the person had programmatic access, and check whatever automation scripts referenced those keys.
- Confirm exported or local profile copies are accounted for. If your workflow allows exporting profiles to disk, know which team members had that permission and assume anything they exported is a standing liability until the underlying accounts are re-secured.
- Check permission levels going forward, not just after the fact. AdsPower and Dolphin Anty both support granular sub-user roles that can restrict export and proxy-visibility permissions ahead of time, which limits how much a departing teammate could have taken with them in the first place. Setting those restrictions before someone leaves is cheaper than cleaning up after.
Why this matters more for multi-accounting specifically
For a normal team SaaS tool, a lingering session from a former employee is annoying but contained. For multi-accounting operations, the exposure compounds. If a former teammate still has a working session on an account that’s also being actively used by someone else through the same profile, you can end up with two different IPs, devices, or usage patterns touching what a platform sees as one continuous session. That kind of inconsistency is exactly the sort of signal platforms use to flag accounts for review, independent of anything to do with the anti-detect browser’s fingerprint quality. Access control failures don’t just create a security risk, they create a detection risk on top of it.
Treat workspace removal as the start of an offboarding checklist, not the end of one. The browser tool controls who can open a profile through its own dashboard. It was never going to control what happens to a session, a password, or a proxy credential once that data left the platform.
If you’re comparing how different anti-detect browsers handle team permissions, profile exports, and shared workspace controls, we test these tools hands-on and cover what actually holds up at Anti-Detect Review.
Get new guides and videos first — join the Telegram channel.