Authors: Assistant Professor M. Sivaparvathi, Assistant Professor D. Sravya, Associate Professor CH. Naveen
Abstract: Users may now share information quickly and simply with the use of laptops, cellphones, and QR codes. The security of this widely used technology is in jeopardy due to the increasing number of hacks that target QR codes. Users must have trust in the QR Code system for transactions to be secure, regardless of whether they are at a conference, restaurant, or payment station. Since 2020, there has been a 2400% rise in the deployment of QR vector malware assaults, and a 50% increase in phishing methods that use bogus codes. The need to restore integrity safeguards in order to regain the public's confidence is emphasised by these concerning outcomes. In response to these worries, this study employs AI and third-party services to identify QR Code security threats. Here we provide a Secure QR Codes system that employs four ML models, including ones for anomaly detection and risk warning, as well as ones trained on good and poor QR Codes. With the help of encryption engines, APIs, proxy services, AI evaluation modules, and predictions based on AI, the integrated system detected more than 96% of the simulated real-world threats, such as phishing, exfiltration, injection, and code execution. We have developed a method that can detect altered QR codes. As a viable, non-proprietary, web-deployable solution to anticipate and minimise these issues, it helps to rebuild confidence in QR Codes systems in many contexts. This index is associated with many subjects, including safe QR code infrastructure, fast response codes, cybersecurity, and anomaly detection.
