Artificial intelligence-guided design of high-entropy materials from bulk to nanoscale

High Entropy Materials (HEMs) have provided a breakthrough in materials science, moving beyond single-principal-element design to multielement (>5) design with multiple elements in near-equiatomic ratios. This compositional complexity introduces high configurational entropy, stabilizing single solid-solution phases and providing outstanding thermal, mechanical, and functional characteristics to alloys, oxides, carbides and nitrides. Although traditional empirical methods are the cornerstone of materials design, they are less efficient in exploring the vast compositional and processing space in the sphere of HEMs. The recent development of artificial intelligence (AI) and machine learning (ML) has altered the field of HEM research by allowing data-driven prediction of phase stability, structure-property relationships, and processing results, which has tremendously accelerated the process of materials discovery. With the combination of thermodynamic descriptors, including, but not limited to, ΔS _mix, mixing enthalpy (ΔH _mix), atomic size mismatch (δ), phase purity (Ω), and valence electron concentration (VEC), AI models enable researchers to perform composition and property optimization more easily than ever before. The AI-guided frameworks, despite their limitations like the lack of data, can enable autonomous and high-throughput design of HEMs. This viewpoint reflects recent developments, examples of energy and catalytic processes, and the future outlook and the potential transformative nature of AI to provide next-generation high performance entropy-stabilized materials.