Next-generation antiviral peptides: AI-driven design, translational delivery platforms, and future therapeutic directions

Antiviral peptides (AVPs) are emerging as next-generation therapeutics due to their broad-spectrum activity, low toxicity, and ability to overcome drug resistance. The objective of this review is to provide an integrated perspective on AVP research, with particular emphasis on artificial intelligence (AI)-driven discovery, novel delivery strategies, and translational applications. We first summarize the origins, mechanisms, and structural diversity of AVPs. We then highlight recent advances in computational pipelines, including machine learning, deep learning, generative adversarial networks (GANs), large language models (LLMs), and reinforcement learning frameworks for de novo peptide design. Translational aspects are addressed by discussing novel delivery systems such as nanoparticles, hydrogels, and intranasal/inhalable formulations, as well as clinical trial examples (like, enfuvirtide (T-20), sifuvirtide, lactoferrin-based formulations, PAC-113). Finally, we explore future directions, including CRISPR- and mRNA-based peptide delivery and synergies with immune checkpoint inhibitors. By combining classical mechanisms with AI-driven design and innovative delivery platforms, this review underscores the potential of AVPs as versatile antiviral agents ready for clinical translation.