Bots are now able to crack Captcha tests faster than humans, according to a new study by researchers at the University of California, Irvine.
Engaging 1,000 participants with diverse backgrounds, researchers conducted 10 Captcha tests on these websites. The results showcased that while human participants took 9 to 15 seconds to solve certain tests with 50-84% accuracy, bots cracked the same tests in less than a second with near-perfect precision.
The study’s findings emphasize the bots’ accuracy surpassing 85-100%, in stark contrast to the human accuracy range of 50-85%. Furthermore, the bots exhibited solving times either “significantly lower” or nearly equivalent to humans in most instances.
Captcha tests are designed to be difficult for bots to solve but easy for humans. They typically involve tasks such as identifying distorted text or selecting images that contain specific objects.
The researchers found that bots are now able to use machine learning to solve Captcha tests with ease. This is a major security concern, as it means that bots can now easily access websites and services that are protected by Captcha.
The researchers called for the development of new Captcha tests that are more difficult for bots to solve. They also suggested that websites should use other security measures, such as IP address restrictions and two-factor authentication, to protect themselves from bots.
The sources for this piece include an article in Independent.