20 Jun 2015
The purpose of a CAPTCHA is simple: protect a website from malicious attacks (i.e. spammers) by being difficult/impossible for bots but easy enough to let humans through. But what happens when the most commonly used CAPTCHA service can be solved with 97%+ accuracy by the very bots it was designed to beat?
For over a decade, text based CAPTCHAs have been the popular choice for this task. They grab a word (usually English), warp it into a shape not commonly seen and then ask users to type the words they see. Some text CAPTCHAs even use a random assortment of letters and numbers in an attempt to hinder the bots even more. The issue? Programs that utilize Optical Recognition Software, known as OCR, read the distorted text and allow bots through to websites that relied on the security service to prevent that very thing happening.
This, unfortunately, is a common problem. By design, text CAPTCHAs have a shelf life – in order for them to remain difficult for bots, they have to become increasingly harder for humans. It appears that we’ve reached the ceiling for text CAPTCHA effectiveness, which is a big motivation for our creation of FunCaptcha.
The internet was built on innovation and that’s exactly what we’re doing with FunCaptcha – innovating an area of web security that sorely needs it.
Update: watch co-founder and CAPTCHA expert, Matthew Ford, go into detail on this topic in our new video series!