AI ethics for earth sciences

Abstract

Following standards for artificial intelligence (AI) ethics and best practices can ensure that researchers and practitioners share better insights across subdisciplines. In the Earth sciences, the practice of responsible AI is critical because it provides our society with a sense of security and value in our daily lives. Responsible AI involves developing a formal ethical framework for AI; however, an exhaustive treatment for all applied AI research, including the earth sciences, has not been developed yet. Nonetheless, some basic best practices and standards are available now, and others are coming out in the short term. This chapter briefly describes the basic standards for AI ethics and best practices that are readily available now. These go from ethical considerations of autonomous AI systems such as those based on satellite data or sensor networks, up to discussing practical guides to document machine learning models. We believe that these principles provide an essential roadmap for what is vital for responsible Earth AI development and use.