Organic Materials of Tomorrow: Horizons of Artificial Intelligence

Artificial intelligence (AI) is transforming organic materials discovery by enabling the rapid exploration of chemical space. This review examines machine learning techniques being used to accelerate the identification of novel compounds for organic semiconductors through computational approaches linking molecular structure to properties. Key methodologies include graph neural networks, generative approaches, chemical representations, $\Delta$ -learning frameworks, machine learning force fields, active learning, transfer learning, and generative models. These methods address fundamental challenges in organic materials discovery, from property prediction and inverse design to high-throughput screening and molecular generation. An example of applications to the topic of organic photovoltaics demonstrates practical impact in predicting energy levels, morphology, charge transport, exciton dynamics, and power conversion efficiency. Rather than replacing human scientists, we envision AI as a tool that amplifies their capacity to explore unconventional regions of chemical space. Advantages, drawbacks and bottlenecks of AI use in chemistry are discussed together with future research directions, such as the adoption of human-centered AI practices, the construction of materials-science-oriented benchmarking databases and protocols, the integration of green chemistry constraints into generative pipelines, and the further exploration of end-to-end in-silico-to-technical validation workflows, all tailored to the needs of the materials science community.