Connect with us

Artificial Intelligence

AI can now breach Anti-Bot Security by thinking like Humans



Artificial Intelligence Failing Anti-Bot Security Systems

Artificial Intelligence Can Now Breach Through CAPTCHA Security?

Few years back when websites were suffering from different kind of bot attacks, the Anti-Bot Security Systems were developed. These systems were quite basic at start like they used to ask “what’s 1+3” at different steps of a website procedure but later these systems developed with time to make the security strong.

All of you must be familiar with the word CAPTCHAs (yes those frustrating online tests which asks you to identify fuzzy numbers and letters). Bots are not able to pass these CAPTCHAs and helps us to prevent any kind of spam. Tech Giants have developed hundreds of types of these CAPTCHAs (Completely Automated Public Turing test to tell Computers and Humans Apart) ranging from Mathematical Reasoning to Visual tests. We are quite sure that no script or bot can be built to bypass this security system, as these anti-bot systems are completely random and requires common sense to answer that only humans posses.

Many developers tried to develop a Anti-Anti-Bot system but failed and stated that there are no flaws in the system and it can’t be breached. But what if I tell you that all of them were working in the wrong direction of research, I mean If these tests requires common sense to pass then they should have worked on building that which is known as Artificial Intelligence these days.

Dileep George (co-founder of AI startup Vicarious) and his team developed a computer model called Recursive Cortical Network (RCN) which can now bypass most of the security CAPTCHAs. George told “Biology has put a scaffolding in our brain that is suitable for working with this world, It makes the brain learn quickly. So we copy those insights from nature and put it in our model.”

In past, many have tried and got success too in bypassing the CAPTCHAs but of a particular type or website only as they were based on training to recognize per-labbeled A,B,Cs while this new AI is based on the idea of deep learning. RCN uses algorithms that enable it to generalize – detecting patterns in contours and surfaces, and able to separate characters even when they overlap.

George tells how RCN aprroach works:

“During the training phase, it builds internal models of the letters that it is exposed to. So if you expose it to As and Bs and different characters, it will build its own internal model of what those characters are supposed to look like. So it would say, these are the contours of the letter, this is the interior of the letter, this is the background, etc. And then, when a new image comes in … it tries to explain that new image, trying to explain all the pixels of that new image in terms of the characters it has seen before. So it will say, this portion of the A is missing because it is behind this B.”

A representation of the letter A, which can be used to crack CAPTCHAs

Representation of letter A in RCN. Credits: Vicarious AI

According to their report RCN was able to solve CAPTCHAs at accuracy rate of 67% including BotDetect at 64%, Yahoo at 57% and Paypal at 57%.

It’s a point to note that the point of this research is not at all to breach security CAPTCHAs, it’s about making robots to think and visualize like human beings. George says “The long-term goal is to build intelligence that works like the human brain, CAPTCHAs were just a natural test for us, because it is a test where you are checking whether your system can work like the brain.”

As now this AI has been developed it’s only a matter of time that these traditional CAPTCHAs will only be found in history, surely these old Anti-Bot Security Systems will be replaced soon. “We’re not seeing attacks on CAPTCHA at the moment, but within three or four months, whatever the researchers have developed will become mainstream, so CAPTCHA’s days are numbered,” security researcher Simon Edwards from tech security firm Trend Micro told the BBC.

George adds “Robots need to understand the world around them and be able to reason with objects and manipulate objects, So those are cases where requiring less training examples and being able to deal with the world in a very flexible way and being able to reason on the fly is very important, and those are the areas that we’re applying it to.”

When asked about the people who feel afraid of the idea of Robots having thinking ability like humans George said that this is going to be a march of technology, we have to start taking it for granted that machines/robots will be able to have same human thinking power soon.

However it’s not clear right now that how much of an impact this research will have on IT Security but it’s right to assume that with the advancement of Artificial Intelligence, more advanced tests will be required to prove that you are a human. This is the reason why Google has already moved from text based CAPTCHAs to more advanced tests. The idea of this research may be fearsome to some but George says it’s all for the good, let us know in the comment section below what you think about this Recursive Cortical Network (RCN) research and advancement of AI.

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Recent Posts