"This New Kind of AI Can Breach Anti-Bot Security by Thinking Like a Human"

With the use of a new type of neural network by the name of Recursive Cortical Network (RCN), the Completely Automated Public Turing test to tell Computers and Humans Apart (CAPTCHA) can now be defeated. CAPTCHA is a security system that is widely used on the internet to protect websites against automated bots by generating online tests in which the user is asked to identify distorted letters and numbers. The objective of CATCHA is to differentiate between humans and automated software robots, which were believed to be unable to detect such obfuscated text until now. The RCN has been developed by researchers to distinguish letters with the use of algorithms to detect patterns in a way that allows for generalization. This article further discusses the traditional concept of CAPTCHA, reasons as to why it has been difficult for AI systems to overcome the barriers of CAPTCHA, how the RCN defeats CAPTCHA, and the ultimate goal of the RCN. 

ScienceAlert reports "This New Kind of AI Can Breach Anti-Bot Security by Thinking Like a Human"