Organizers launched the AMP Challenge at NeurIPS 2026, introducing a standardized, blinded benchmark for generative AI-designed antimicrobial peptides against clinical bacterial pathogens. The initiative is intended to provide consistent evaluation criteria across models and reduce variability in how peptide generation systems are assessed. Benchmarking is particularly consequential for antimicrobial peptide programs because in vitro hits often fail to translate without robust design constraints and evaluation consistency. A blinded test can help prevent overfitting to informal scoring rubrics. For AI drug discovery teams, the challenge signals a shift toward shared evaluation infrastructure for antimicrobial peptide design claims.