This New AI-Based Tool Will Prevent Cyber Attacks in Websites

Now the world’s popular websites can prevent cyber-attacks with the new AI-based tool. A team of researchers from three different countries, i.e. China, the US and the UK has developed an AI-based toll which can read text captcha schemes which have the potential to safeguard the websites from being hacked or breached.

The new technology is based on deep learning methods and till now it is one of the most effective solver of captcha security and authentication systems. Created jointly by a group of researchers from Lancaster University, Northwest University and Peking University, the tool is capable of cracking versions of captcha where attacks systems have previously failed to do so.

According to the researchers, the new tool can solve a captcha within a quick span of time, i.e. within 0.05 seconds. The mechanism involved in the process is – a captcha generator which is taught to produce large numbers of captchas that are indistinguishable from genuine captchas. 

The researchers used a machine learning automatic captcha generator and were able to significantly cut down the time and efforts which required to search and conduct the manual tagging of the captchas to train the software.

So, the new tool requires very less human involvement and it can target new, or modified, captcha schemes with ease.

The new tool was tested on 33 different captcha schemes. Some of them are used by Microsoft, Wikipedia and eBay.