The National Academy of Medicine (USA), as part of its AI Code of Conduct initiative, has presented a proposed Code of Conduct for the use of AI in healthcare, care delivery, and the biomedical sector.
National Academy of Medicine – Draft Code of conduct for artificial intelligence in health, health care and biomedical
Anno 2024
This proposed Code of Conduct aims to provide a set of guidelines for the development and application of AI in health, healthcare services, and biomedical science. It identifies the principles that should guide the development and use of AI in the mentioned areas, along with specific commitments to be undertaken for the concrete implementation of those principles.
Specifically, the proposed Code of Conduct outlines the following ten principles:
- Participation: It is necessary to understand, express, and prioritize the needs, preferences, and goals of individuals, as well as their implications, throughout the entire lifecycle of AI.
- Safety: Attention and continuous vigilance must be ensured regarding potentially harmful consequences for individuals and population groups that may arise from the application of AI in health and medicine.
- Effectiveness: The application of AI must have demonstrated its ability to achieve the intended improvement in personal health and human condition, in compliance with established ethical principles.
- Equity: The request to employ AI in the specified sectors must be accompanied by evidence of appropriate measures adopted to ensure fair and impartial development of this technology, equitable access to AI’s benefits, and risk mitigation measures.
- Efficiency: It must be ensured that the development and use of AI is associated with a reduction in healthcare costs, as well as a decrease—or at least the maintenance—of current negative environmental impacts.
- Accessibility: It is necessary to ensure that seamless access and involvement of stakeholders is a fundamental feature at every stage of the AI lifecycle and governance.
- Transparency: Open, accessible, and understandable information must be provided regarding the components of the AI system, its functioning, and its outcomes.
- Accountability: Actions taken during the development and use of AI must be identifiable and measurable, with clear documentation of benefits and resulting forms of responsibility concerning potential adverse consequences.
- Protection: Procedures and validation mechanisms must be adopted to ensure privacy and security protection, since health data sources must be considered elements to protect more effectively for the pursuit of the common good, even when AI use is functional to learning and continuous improvement.
- Adaptability: It is essential to ensure that the framing of responsibility levels continuously provides information on the outcomes produced by AI systems concerning their use and improvement in medicine, healthcare, biomedical science, and, ultimately, the human condition in general.
With regard to the commitments to be undertaken to develop and foster trust in all decisions concerning the development and dissemination of AI systems, the proposed Code of Conduct identifies six main commitments:
- Make the protection and improvement of human health and the strengthening of human connection primary goals;
- Ensure a fair distribution of the benefits and risks associated with the use of AI;
- Engage individuals as active partners at every stage of the AI system’s lifecycle;
- Reinforce in healthcare professionals a sense of moral well-being and a shared purpose;
- Monitor and openly share, in a clear and understandable manner, the methodologies and evidence related to the functioning of AI and its impact on health and safety;
- Innovate, collaborate in learning, and continuously improve the standards of clinical practice.
The proposed Code of Conduct will be open for public consultation until May 3, 2024.
The full text of the document is available at the following link and in the download box.
Maria Vittoria Izzi
Pubblicato il: Lunedì, 08 Aprile 2024 - Ultima modifica: Lunedì, 01 Settembre 2025
