Hackers found a way around MFA entirely, and it’s already led to ransomware attacks
Security researchers say they’ve uncovered a hacking campaign that manages to sidestep one of the most trusted defences in cybersecurity, multi-factor authentication (MFA), by stealing sessions from people who are already logged in, rather than trying to guess their passwords.
The findings come from Huntress, a cybersecurity firm that investigated six separate break-ins at unrelated organisations earlier this year. In each case, the attackers used the same method to get in, the same trick to gain full control of the systems they landed on, and, in the most serious case, the same end goal: deploying ransomware from a group called DragonForce.
How does an attack get around MFA?
The attacks exploit a flaw, nicknamed “CitrixBleed 2”, in Citrix login systems that many businesses use to let staff work remotely. Sending the system a slightly broken login request causes it to accidentally leak small snippets of its own memory back in the response. Sent enough times, those snippets start to include session tokens, essentially digital proof that a real employee is already logged in.
Steal one of those tokens, and an attacker can effectively become that employee without ever entering a password or MFA code, because the system believes they’re already authenticated. In one case Huntress investigated, researchers could see a legitimate staff member log in normally, and just 21 minutes later, watch the same session being used from the attacker’s own internet address with no separate login ever taking place.
A worryingly consistent playbook
What struck researchers most wasn’t just the technique, but how identical each incident was. The same fake user accounts, the same computer name appearing on the attacker’s end, and the same follow-up steps recurred across entirely unrelated companies. That kind of consistency, Huntress says, suggests the attackers are running a well-rehearsed, repeatable script, likely a group that specialises in breaking in and then handing off (or selling) that access to others, rather than a one-off opportunist.
Once inside, the attackers use a clever trick involving a normally obscure Windows feature to persuade the system to grant them full administrator rights, then quietly reset everything to make it look as if nothing happened. From there, they install everyday remote-access tools, the same kind IT support teams use legitimately, to maintain access without raising alarms.
Where it ends: ransomware
In the most advanced case researchers tracked, the intrusion escalated to DragonForce ransomware, a group known for extorting businesses by encrypting their systems and demanding payment. Huntress says its team managed to contain the damage to a single machine before it spread further, but the intent was clear.
What businesses should do
Huntress is advising any organisation using the affected Citrix systems to patch immediately, and warns that simply patching isn’t enough on its own. Stolen login sessions can remain valid even after an update, so businesses are also being told to log out all active sessions, keep longer-term copies of their system logs, and check for suspicious accounts or remote-access software they don’t recognise.