Many safety primitives are based on tough mathematical issues. The usage of tough AI issues for protection is emerging as an exciting new paradigm, however has been under-explored. In this project, we present a new protection primitive primarily based on hard AI issues, specifically, a family of graphical password structures constructed on top of Captcha technology, which we call Captcha as graphical passwords (CaRP). CaRP is both a Captcha and a graphical password scheme. CaRP addresses a number of security issues altogether, such as online guessing attacks, relay attacks, and, if combined with dual-view technologies, shoulder-surfing attacks. Considerably, a CaRP password may be found handiest probabilistically by automated online guessing attacks even though the password is within the search set. CaRP also offers a unique method to deal with the well-known photo hotspot problem in popular graphical password structures, such as PassPoints, that regularly leads to susceptible password selections. CaRP isn’t always a panacea, but it gives reasonable protection and usability and appears to fit properly with some practical applications for enhancing online protection.