INTEGRITY GAURD: An AI-Driven Proctoring System for Fair Online Examinations

The rapid rise of online education in the context of the pandemic, and following it, has introduced a remarkable

change in testing methodology from onsite conventional exams to digital platforms . This shift has indeed offered increased

accessibility and scalability, however, there are concerns for fairness, academic integrity and the trust issues . Studies have

found that a large number of students have admitted cheating in these online tests, which makes the trust level of online

examination lower compared to traditional examination method . Technical problems like unreliable connections and security

concerns. In response to these challenges, this work suggests an AI-enhanced remote proctoring framework, which combines

multiple modes of monitoring. The design base is essentially: facial recognition for person identification and intruder detection

audio analysis for background conversations detection (it includes references ) and behavioral monitoring of facial gazing, head

posture, and eye track movements. Furthermore, monitoring screen and tab activity might raise a red flag that something fishy is

going on in the digital life Line. Dynamic cheating score measures abnormal behavior and produces automated logs, assisting in

decision making of examiners. Acknowledging ethical issues, privacy protection, encryption procedures, and transparent policies

as part of the framework in order to alleviate student worries and meet data protection requirements . Fairness-aware AI models

implemented to mitigate bias amongst different student groups . Arresting the pendulum between innovation and ethics, this

research highlights the promise of sophisticated AI-powered proctoring systems for increasing the credibility, equity and

trustworthiness of online assessment. The solution offers institutions a scalable, trusted solution that upholds the integrity of the

academic process while treating students with respect.