The abbreviation Captcha stands for “Completely Automated Public Turing test to tell Computers and Humans Apart” and is a test to distinguish between humans and computers (bots). Captchas became necessary to protect operator resources and to make sure it isan actual human entering data into online forms, since bots are often misused (keyword: spam).
In most cases, CAPTCHAS consist of a challenge-response test in which the user must solve a task, the so-called challenge, and the result is sent back in the response. The task is created fully automatically by a random generator, adhering to certain rules.
Classic text-based captchas are designed so that letters and numbers must be identified that are either displayed in distorted form or with additional characters in the background. But there are many other captchas, such as:
- Image-based captchas: The user is usually presented with square images and certain objects such as cars, bridges, zebra crossings, etc. must be recognized and selected.
- Mathematical captchas: A mathematical problem is given that the user has to solve. However, since computers can solve mathematical problems very easily, reading the problem is made difficult by the background used, or the problem is more of a puzzle and, for example, only the first digit of the result has to be entered.
- Audio-based CAPTCHAS: Since text or image CAPTCHAS can only be solved if the user has the appropriate visual abilities, some CAPTCHAS are used in combination with an audio task. The user is presented with an audio file and the correct term must be typed in. Often, a lot of background noise is used to prevent the spoken text from being automatically transcribed and entered.
- Interactive CAPTCHAS: These range from simple tasks such as clicking on a particular object or checkbox, to complex tasks such as correctly aligning an object based on the given direction.
In response to CAPTCHAS, Google launched reCaptcha. The idea behind it is the same as for CAPTCHAS: humans should be able to solve it easily, but not computers/bots. In the early days of ReCaptchas, the service also helped Google to identify and digitize text. The user was shown a word that could not be digitized by the optical recognition software and the user had to complete this text. With this help, Google was also able to improve Street View by helping users to identify house numbers and street names.
In 2013, behavioral analysis was added to examine users' browser interactions and calculate a probability of whether the user is a human. If this is the case, the now famous “I'm not a robot” field is displayed, which must be confirmed with a mouse click. If, on the other hand, the user cannot be identified beyond doubt as a human, they have to solve a captcha.
Captchas often massively restrict usability, as a captcha test can interrupt the flow of a user's activities or be almost impossible to solve for visually impaired people. Also, computers are getting better and better at solving captchas, while it is becoming increasingly difficult for humans, which is why captchas will not be a long-term solution.