This study investigates the response of institutions to cases of high AI and high similarity in higher education settings by evaluating global best practices and current challenges. It seeks to identify effective mechanisms, policies, and strategies that respond to the increasing development of AI-generated content and implications related to academic integrity. The literature review was undertaken in a systematic way according to the PRISMA structure to ensure transparency and rigor. Qualitative thematic analysis was performed using NVivo software, which enabled the identification of key trends and gaps in institutional practices. Key findings reveal a diverse range of strategies, including integrating academic integrity education into curricula, adopting AI detection tools like Turnitin and SafeAssign, and implementing tiered penalty systems for proportional responses. Other challenges include false positives, ethical issues, and gaps in policies that highlight the use of balanced approaches between prevention, detection, and education. The study concludes by mentioning that the challenges indeed call for global collaboration and innovation. Institutions should have explicit and adaptable policies, investment in reliable AI detection technologies, and foster transparency with ethics in AI use. These steps are crucial in ensuring academic integrity in the modern educational space that is being shaped by AI developments.