SecureGenKeys: Building Secure and Personalized Keypads with Generative AI

Ensuring the security of systems is paramount in safeguarding confidential information from unauthorized access, particularly in today's interconnected digital landscape. Visual eavesdropping techniques, such as “Shoulder Surfing,” pose significant threats to security, including unauthorized access, data breaches, and identity theft, whereby sensitive information is illicitly obtained during user input. This paper introduces a novel method designed to fortify password security, simultaneously enhancing the user experience and facilitating secure transactions. The proposed methodology introduces a personalized interactive password selection process that not only increases user engagement but also bolsters security measures. User-entered prompts undergo processing to transform into personalized tokens and form a secure password. These tokens are utilized to generate foundational images in which the tokens are strategically allocated ensuring both clarity and randomness. The system employs machine learning algorithm, particularly focusing on deep learning approaches for image analysis and pattern recognition techniques to verify user-entered passwords against stored data. Successful authentication triggers secure transactions, providing a seamless and protected user experience.