Determining the ethical codes of artificial intelligence-assisted informed consent

Title: Determining the Ethical Codes of Artificial Intelligence-Assisted Informed Consent Aim: This doctoral thesis aims to determine the ethical codes for Artificial Intelligence (AI)-assisted informed consent systems by analyzing the ethical issues that may arise in their implementation. Method: The study involves a comprehensive review of national and international guidelines and regulations concerning informed consent and AI ethics. Additionally, advanced AI programs like GPT-4 based ChatGPT were presented with informed consent scenarios to assess their ethical behavior, and the obtained data was analyzed in light of traditional ethical principles and AI ethics guidelines. Results: While traditional ethical principles retain their importance in the age of AI, new principles such as explainability, accountability, and auditability emerge as central. AI systems hold the potential to mitigate issues arising from language and cultural differences, provide more detailed information, and minimize potential biases, yet the lack of human elements like empathy, compassion, and mourning remains a concern. Conclusion: The study proposes 21 ethical codes to be considered in the implementation of AI-assisted informed consent systems. These codes aim to ensure that AI systems are transparent, accountable, and just while protecting patient autonomy, privacy, and safety. Furthermore, highlighting the importance of human oversight, a novel ethical approach termed "metadeontic" is proposed. This approach envisions the adaptation of universal and immutable ethical principles to individual needs and cultural differences through AI, effectively individualizing them within the bounds of universality.