The rapid integration of artificial intelligence (AI) into higher education has revolutionised academic research and teaching, offered transformative opportunities while raising significant ethical challenges. This Delphi study investigates the ethical dilemmas and institutional requirements for maintaining academic integrity in AI-driven research. Through a structured, iterative process involving 12 international experts from academia, AI regulatory bodies, and research ethics committees, the study identifies key concerns such as data privacy, algorithmic bias, transparency, and authorship in AI-assisted research. The findings reveal a pressing need for clear ethical guidelines, robust institutional oversight, and enhanced collaboration between academia and industry to ensure responsible AI use. Experts emphasise the importance of developing enforceable policies, integrating AI ethics into academic curricula, and fostering transparency in AI decision-making processes. The study concludes that while AI offers immense potential for innovation, its ethical implications demand proactive governance and continuous policy refinement. By addressing these challenges, higher education institutions can uphold academic integrity, mitigate risks of misconduct, and promote the responsible development and application of AI technologies. This research provides actionable recommendations for policymakers, educators, and researchers, contributing to the creation of a sustainable and ethically sound AI ecosystem in academia.