Article 6: Transparent and explainable.
Continuously improve the transparency of artificial intelligence systems. Regarding system decision making processes, data structures, and the intent of system developers and technological implementers: be capable of accurate description, monitoring, and reproduction; and realize explainability, predictability, traceability, and verifiability (可解释、可预测、可追溯和可验证) for algorithmic logic, system decisions, and action outcomes.
IEEE supports the inclusion of ethical considerations in the design and deployment of autonomous and intelligent systems.
Autonomous and Intelligent systems (A IS) are systems that are capable of adaption and learning based on feedback and data from their environment. A IS hold great promise to benefit society in applications domains as diverse as transportation, health and social care, environmental preservation, enterprise productivity, communication network optimization, power grid adaptation and management, agriculture, manufacturing, and entertainment. Recent success in machine learning, signal processing, planning algorithms, digital sensing, embedded systems, cloud computing, as well as voice, image and pattern analysis have greatly accelerated application of A IS. They hold great promise to benefit society, but they also present potential new social, legal and ethical challenges, with corresponding new requirements to address issues of systemic risk, diminishing trust, privacy challenges and issues of data transparency, ownership and agency.
Therefore, there is a compelling need for developers and operators of A IS systems to maintain awareness of and employ consensus based global best technical practices and standards that recognize and align end users’ and citizen’s values when building and deploying A IS. To that end:
3. Clear responsibility: The development of artificial intelligence should establish a complete framework of safety responsibility, and we need to innovate laws, regulations and ethical norms for the application of artificial intelligence, and clarify the mechanism of identification and sharing of safety responsibility of artificial intelligence.
Artificial intelligence should be safe and reliable. We are dedicated to accentuating technical robustness and security throughout the research process, providing a secure and reliable system to improve the ability to prevent attack and conduct self repair.
Artificial intelligence should be transparent and interpretable. We are committed to conducting open source and interpretative research, reducing research on blind black box algorithm, and enhancing multi layered transparency, thus attesting to the compliance with the proposed framework of ethics.