Everyone has heard “don’t reuse passwords”. The reason is a specific attack called credential stuffing: criminals take usernames and passwords stolen from one website and try them, automatically, on everything else. If one of your staff uses the same password for a shopping site, their Gmail, the income-tax portal and the office server, a breach at the shopping site can open all four.
How credential stuffing works
A website is breached and its list of email addresses and passwords leaks. Attackers do not stop at that one site. They load the list into tools that try every pair against email services, banking sites, cloud apps and remote-access logins such as Remote Desktop. Nobody has to guess anything; they are simply testing whether people reused a password that is already known.
The tools run around the clock, spread their attempts across many addresses so one blocked address does not stop them, and try a few passwords per account at a time to stay under lockout limits. Against a Windows server, that shows up as failed sign-ins (event ID 4625) for many different usernames, often from many different IP addresses. Our post on reading event ID 4625 shows what to look for.
Where the stolen passwords come from
Passwords have leaked from large services, among them LinkedIn, Adobe and Dropbox, and the lists are copied, combined and passed around for years afterwards. You can check whether your own email addresses appear in known breaches at Have I Been Pwned, a free and widely used service. It is worth running every partner’s and staff member’s work and personal email through it once; an address that shows up there should have its password changed everywhere it was used.
What reuse looks like in a CA office
In a small Indian firm, reuse usually comes from convenience during a busy filing season. A typical pattern:
- The same password on the income-tax e-filing portal, the GST portal and the MCA login for a client, often shared among two or three staff.
- The staff member’s personal Gmail or a shopping app using the same password as their Windows login on the office server.
- The Tally login, the server login and the Wi-Fi password all built from the firm’s name and the year, such as “Firm@2026”. This is an example of what not to do; it is near the top of every guessing list.
One leak anywhere in that chain gives an attacker a password that works on the server, where your clients’ PAN, bank and GST data sit.
How to protect the firm
Use a different password for every account. This is the defence that breaks credential stuffing: a leaked password is useless anywhere else. A password manager such as Bitwarden, 1Password or Keeper makes it practical, and lets partners share portal logins with staff without everyone knowing the same password.
Make server passwords long. We recommend at least 15 characters, in line with NIST’s 2025 guidance; see how to set the minimum password length on Windows Server.
Turn on multi-factor authentication (MFA) wherever it is offered. With MFA, a stolen password on its own is not enough to sign in, which defeats ordinary credential stuffing. It is not a perfect shield. Adversary-in-the-middle phishing kits put a fake login page between the user and the real site and pass the one-time code through as it is typed, so SMS and app codes can be stolen too. That is why the US cybersecurity agency CISA recommends phishing-resistant MFA such as security keys or passkeys for the most important accounts. Our guide to two-factor authentication covers the options.
Getting MFA onto Remote Desktop
Windows Remote Desktop has no built-in MFA, so “turn on MFA for remote access” needs one of these:
- Guard Secure Access. Staff reach the server through an encrypted tunnel, and Remote Desktop is closed to the internet. Only devices you have invited can connect, so a leaked password on its own does not reach the server. With the default “Prixl Magic” connection method it needs no router setup and no static IP.
- Remote Desktop Gateway with MFA, for example Microsoft’s NPS extension for Microsoft Entra MFA. It works well but needs extra server roles and Microsoft licences; it is a job for your IT vendor.
- A third-party MFA agent installed on the server that asks for a second factor at Windows sign-in.
Whichever you choose, the first step is the same: stop Remote Desktop from answering the whole internet. How to restrict RDP access safely walks through it.
Watch for it happening
Even with good passwords, you want to know when someone is testing them. Prixl Warden reads the Windows sign-in logs on your server and emails you when it sees brute-force patterns, per IP address and per account, whether they come as fast floods or slow, low-and-slow campaigns. That is how credential stuffing against a server shows up in the logs. You can switch on automatic blocking of attacking IP addresses in its dashboard. For the wider picture of password attacks, see brute-force attacks and how to stop them.

