What are FDA’s Guiding Principles of Good AI Practice in Drug Development? – Healthcare Economist

Here they are:

1. Human-centered by design. The development and use of AI technologies is consistent with ethical and human-centered values.
2. risk based approach. The development and use of AI technologies follows a risk-based approach with proportionate validation, risk mitigation and oversight based on the context of use and determined modeled risk.
3. adherence to standards. AI technologies adhere to relevant legal, ethical, technical, scientific, cybersecurity and regulatory standards, including good practices (GXP).
4. Clear context of use. AI technologies have a well-defined context of use (role and scope for why it is being used).
5. Multidisciplinary expertise. Multidisciplinary expertise covering both the AI ​​technology and the context of its use is integrated across the life cycle of the technology.
6. Data Administration and Documentation.Data source provenance, processing steps and analytical decisions are documented in a detailed, traceable and verifiable manner in line with GxP requirements. Appropriate governance, including privacy and security for sensitive data, is maintained throughout the life cycle of the technology.
7. Model Design and Development Practices. The development of AI technologies follows best practices in model and system design and software engineering and leverages data appropriate for use, considering interpretability, explainability, and predictive performance. Good model and system development promotes transparency, reliability, generalizability, and robustness for AI technologies that contribute to patient safety.
8. Risk-Based Performance Appraisal. Risk-based performance assessments evaluate the entire system, including human-AI interactions, using appropriate data and metrics appropriate for the intended context of use, supported by validation of predicted performance through appropriately designed test and evaluation methods.
9. life cycle management. Risk-based quality management systems are implemented throughout the life cycle of AI technologies, including helping to capture, evaluate, and address issues. AI technologies undergo scheduled monitoring and periodic reassessment to ensure adequate performance (for example, to address data drift).
10. Clear, essential information. Simple language is used to present clear, accessible and contextually relevant information to the intended audience, including users and patients, regarding the use, performance, limitations, underlying data, updates and interpretation or clarification of AI technology.

You can read the full document here.