Researchers unveiled DeepTarget, a computational platform that predicts mechanisms of action for small molecules, and separate work from Cellarity described an AI‑driven framework for predicting drug‑induced liver injury (DILI) using integrated multi‑omics. Both advances were published recently and illustrate a surge in AI‑led tools aimed at accelerating target ID and de‑risking safety liabilities. DeepTarget (Sanford Burnham Prebys, npj Precision Oncology) leverages large‑scale computation to suggest anti‑cancer mechanisms beyond one‑drug/one‑target assumptions, potentially informing hit‑to‑lead and repurposing strategies. Cellarity’s approach combines cell‑state models with multi‑omics and machine learning to forecast hepatotoxicity signals earlier in preclinical pipelines. These platforms aim to reduce late‑stage failures by improving mechanistic understanding and toxicity prediction; however, adoption will hinge on prospective validation and integration with experimental workflows. For biopharma R&D, the tools underscore a pivot toward computationally guided decision making across discovery and safety assessment.