What are MFA attacks and how can they be prevented?

Multifactor authentication (MFA) is a security measure that requires users to provide a second form of verification before they can log into a corporate network. It has long been considered essential for keeping fraudsters out.
However, cybercriminals have been discovering increasingly clever ways to bypass it.
During an attack on Uber’s IT systems in 2022, the hackers did not use any sophisticated tactics to gain access. Instead, they bombarded an employee with repeated login requests until, out of sheer frustration, the employee approved one.
This type of cyberattack is known as an “MFA fatigue attack” and poses a real risk to organisations, says Anna Collard, SVP of Content Strategy and Evangelist at KnowBe4 AFRICA, a cybersecurity training designer.
“MFA fatigue attacks, also known as prompt spamming or authentication bombing, exploit human vulnerability, rather than relying on high-tech hacking methods,” she explains. “These attacks involve sending continuous push notifications to a target who has already provided their username and password, aiming to irritate or confuse them into unwittingly granting the attacker access to their account or system.”
Previously, cybersecurity experts believed that MFA was a foolproof method to protect corporate IT systems from hackers. “Now we’re seeing attackers finding ways around it by bombarding the victim with scores of MFA requests or tricking them over the phone,” says Collard.
This tactic, similar to a swarm of bees overwhelming someone, is a simple yet effective social engineering technique used by hackers. “By bugging you repeatedly until you give in, malicious actors can manipulate users into approving fraudulent access attempts,” says Collard.
How can you prevent it
The best way to prevent MFA fatigue attacks in organisations is not to use push notifications. “While MFA provides an extra layer of security, it’s not foolproof,” she asserts. “From a cybersecurity perspective, I would recommend that organisations disable push notifications altogether and rather use alternative verification methods.”
An example of a better verification method is number matching. “This involves matching a unique code provided by the authentication app with the code displayed on the screen during the login process,” explains Collard.
A challenge-response method is another effective way of providing additional security. This method asks a user a specific question to verify their identity or to perform a task in response to a challenge. “A challenge-response method is more difficult for hackers to bypass. It can involve mechanisms like biometric authentication, in which users must scan their fingerprints or irises or use facial recognition to gain access to a network.”
However, both of the above are not immune against so-called man-in-the-middle or social engineering attacks tricking the users into handing over their OTP or response to the fraudster.
Another effective verification method is FIDO2, an open authentication standard that allows users to log in without using passwords. “You can implement FIDO2 using hardware security keys,” she explains. Typically, USB sticks store the user’s private key, while the public key is stored on the authentication server. As soon as the user enters their username and password, the system requests them to use the hardware key. “It is more resistant to phishing as it works on a challenge-response protocol and doesn’t rely on a one-time PIN that can be intercepted,” she adds.
Mindfulness is key
As with all hacking attempts, users must remain calm and mindful, rather than reacting emotionally. “Stay tuned into your body’s responses when dealing with potential cybersecurity threats, whether they are phishing emails or MFA fatigue attacks,” says Collard. “If something feels strange, like if the situation is putting you under undue pressure, listen to that cue and don’t respond in a knee-jerk fashion. In this way, you’ll keep a straight head and thwart potential data breaches.” – Distributed by APO Group on behalf of KnowBe4.