De-extinction and beyond: trait design powered by generative AI.

Abstract

Functional de-extinction is the process of creating organisms with traits resembling those of extinct species that benefit modern ecosystems. Advances in genome-wide sequencing, gene editing and traditional cloning technologies have substantially accelerated progress toward de-extinction. However, beyond ongoing ethical debates, this field still faces significant technical challenges. In this opinion, we outline current de-extinction strategies and discuss their advantages and limitations. We then propose a forward-looking framework for de novo trait design by integrating generative AI, including genomic language models, together with ex vivo life engineering. This AI-driven approach to de-extinction may offer powerful tools for uncovering developmental regulatory mechanisms and promoting biodiversity conservation.