<p>The use of artificial intelligence (AI) in endoscopic diagnostics is currently enhancing the accuracy and efficiency of detecting conditions such as pre-cancerous lesions and of other gastrointestinal disorders. AI algorithms, including computer-aided detection (CADe) and computer-aided diagnosis (CADx), help physicians by providing real-time diagnostic suggestions. These technologies improve diagnostic accuracy, especially for conditions with traditionally low sensitivity. However, the integration of AI alongside human clinicians in dual diagnostic systems raises profound ethical questions particularly regarding informed consent. A critical challenge lies in the ‘black-box’ nature of AI, which often obscures how algorithms generate conclusions. This opacity can lead to patient confusion, especially when AI-generated diagnoses differ from those of clinicians. Without clear communication about AI’s role, limitations and potential biases, informed consent may be rendered incomplete thus undermining patient trust and autonomy. This article advances the existing literature by proposing an AI-Specific Layered Consent Model for endoscopic practice. Rather than reiterating general transparency obligations, the layered consent operationalises informed consent by mapping distinct consent duties to the level of AI influence, epistemic opacity and clinical consequence at each stage of the endoscopic pathway. Building on contemporary work in explainable AI, relational autonomy and shared decision-making, the paper develops concrete and stage-sensitive consent obligations tailored to AI enabled endoscopy<i>.</i> The article then concludes with actionable recommendations for improving informed consent processes, emphasising the need for tailored communication strategies that clarify AI’s contributions and limitations. Such measures are essential to exploring ethical challenges in emerging technologies while safeguarding patient autonomy and trust.</p>