If an application has insufficient anti-automation checks in place, attackers can simply keep guessing at passwords until they find a match. Heres how to stop them.
Imagine going to the door of an old speakeasy or underground club. The little hole in the door slides open and a burly bouncer asks for the password. The potential visitor doesn't know the password and makes a guess. It's wrong, so the bouncer doesn't let them inside.
That's what normally would happen. Now imagine the visitor who guessed the wrong password immediately tries again, gets it wrong, and is again denied access. Then imagine the potential visitor opens up the dictionary and starts reading off words, starting with something like aardvark and proceeding to try every single possible word.
Most likely, the bouncer wouldn't allow that kind of activity to take place, but websites and applications with insufficient anti-automation do just that. They allow users to keep trying passwords, even using automation techniques, until they finally stumble across the proper catch phrase.
In this episode, we will learn:
Employing automation or dictionary-style attacks like our imaginary speakeasy visitor did are not new in cybersecurity. In fact, those brute-force style attacks were some of the first hacker techniques ever deployed. And as computers grew faster, they became more and more efficient. A fast computer can run through an entire dictionary of words in just a few minutes, depending on the speed of the connection between the attack computer and the targeted system.
Those kinds of automated attacks were why anti-automation software and techniques were created. It gives applications the ability to determine if actions being taken by a user are outside the norms of typical human behavior.
If an application has insufficient anti-automation checks in place, attackers can simply keep guessing at passwords until they find a match. Or, they might use automation software to do other things such as spam comments into website forums.
Allowing malicious users to employ automation to try and circumvent security can be dangerous. The reason that automation type attacks have persisted from the early days of computing until now is that they can be highly effective. If you give an automation program an unlimited amount of time to submit passwords with no consequences for an incorrect guess, it will eventually find the right one.
When used on something like a forum, having waves of obviously scripted comments might frustrate valid users, or even act like a kind of denial of service attack by squandering system resources. Automated posting might also be used as a tool for a phishing or other attacks to expose the lures to as many people as possible.
To fix the problem of insufficient anti-automation, all applications must be given the ability to determine whether actions being taken are being implemented by a human or a piece of automation software. One of the most popular and widely used techniques is the Completely Automated Public Turing test to tell Computers and Humans Apart, or CAPTCHA.
The CAPTCHA is basically a Turing test, first proposed by computer scientist Alan Turing in 1950, whereby human and computer behavior can be separated and identified. Modern CAPTCHAs present problems humans can easily solve, but which computers struggle with, or simply can't figure out. A popular one presents a photo separated by a grid and asks users to identify all the sectors with a specific item in it, such as a flower or a face. The computer can't understand what is being asked for, and thus can't even attempt to scan the image. Even if it could, image recognition is beyond most programs not specifically built to do so.
Other examples of CAPTCHAs include showing blurry text, asking a simple logic question or even playing the question out loud. Implementing a CAPTCHA challenge at critical points in an application, such as when prompting for a password, can stop automation programs in their tracks.
It's also possible to stop automation programs by simply limiting the number of incorrect guesses from the same source. If too many wrong guesses are sent in, the account can be temporarily locked out, thus delaying the automation program past the point of usefulness, or might even require a human administrator to unlock. Doing any of that should prevent anti-automation vulnerabilities within an application.
For further reading, you can take a look at what OWASP says about insufficient anti-automation. You can also put your newfound defensive knowledge to the test with the free demo of the Secure Code Warrior platform, which trains cybersecurity teams to become the ultimate cyber warriors. To learn more about defeating this vulnerability, and a rogues'gallery of other threats, visit the Secure Code Warrior blog.
Ready to find and fix insufficient anti-automation right now? Test your skills in our game arena: [Start Here]