Dear friends,

This month GCRI announces two new research papers. First, Moral Consideration of Nonhumans in the Ethics of Artificial Intelligence led by GCRI Research Associate Andrea Owe, addresses the current state of treatment of nonhumans across the field of AI ethics. The paper finds limited existing attention and calls for more. The paper speaks to major themes in AI ethics, such as the project of aligning AI to human (or nonhuman) values. Given the profound current and potential future impacts of AI technology, how nonhumans are treated could be of large and potentially catastrophic consequence.

Second, AI Certification: Advancing Ethical Practice by Reducing Information Asymmetries, led by Peter Cihon of GitHub, examines how certification can be used to advance AI governance. Certification can include training certificates for people working in AI and certificates attesting to whether AI systems and the teams that develop them meet certain standards of quality. Certification is used the help outside parties learn what is happening “on the inside” and can play a vital role in good governance of AI. The paper documents existing AI certification programs and discusses how certification may be of value for future AI technology.

These papers demonstrate GCRI’s ongoing commitment to research that bridges the divide between theoretical concepts and practical decision-making for important issues in global catastrophic risk.

Sincerely,

Seth Baum, Executive Director