Abstract
Generative artificial intelligence (AI) is increasingly used in higher and doctoral education for literature exploration, academic writing, coding, data analysis, and access to knowledge. This paper examines how generative AI can support doctoral education while preserving academic integrity, researcher autonomy, and independent scholarly judgement. An integrative literature review was conducted across recent scholarly studies, systematic reviews, editorial and publisher guidance, and institutional policy literature, focusing on AI literacy, cognitive offloading, authorship and disclosure, hallucinated citations, and doctoral supervision. The review indicates that generative AI can improve efficiency, accessibility, feedback, and research support, but unregulated use can encourage over-reliance, weaken critical engagement, obscure authorship and originality, and introduce unverifiable or fabricated information. Based on these findings, the paper proposes the TRUST framework, Transparency, Responsibility, Understanding, Supervision, and Tiered Use,as a conceptual governance model for responsible AI integration in doctoral education. The framework places human accountability and researcher agency at the centre of AI-supported scholarship and proposes differentiated boundaries according to research function and intellectual risk. The paper concludes that doctoral programmes should move beyond blanket prohibition or AI detection alone and instead develop transparent, literacy-based, and supervisory approaches that treat generative AI as an assistive technology subject to human oversight.