EMBL‑EBI, Google DeepMind, NVIDIA and Seoul National University released a vast set of AI‑predicted protein complex structures into the AlphaFold Database. The consortium published millions of models mapping pairwise and multi‑subunit interactions, giving researchers structural hypotheses for protein–protein interfaces previously inaccessible at scale. The dataset is immediately available through EMBL‑EBI and is expected to accelerate target validation, structure‑guided drug design and hypothesis generation across academia and industry. Protein complexes here refer to multi‑chain arrangements of proteins (as opposed to single‑chain structures) and can reveal interaction surfaces useful for small‑molecule or biologic design.