United States
Inductive Bio, a New York–based artificial intelligence company focused on drug discovery, has been awarded up to $21 million in funding from the Advanced Research Projects Agency for Health (ARPA-H) to develop next-generation toxicity prediction models that could reduce reliance on animal testing in pharmaceutical development.
The initiative aims to combine AI-driven modeling with data derived from human tissues and organoids, enabling researchers to better predict drug safety earlier in development. Amgen is participating in the program as an industry partner, contributing data and expertise to support the creation and validation of these models alongside academic collaborators.
The project seeks to address long-standing challenges in preclinical toxicology, where animal models often fail to accurately predict human responses. By integrating machine learning with experimentally generated human biological data, the program is designed to create more relevant and scalable approaches to safety assessment.
Inductive Bio plans to apply its proprietary “virtual chemistry lab” platform to train toxicity models capable of simulating how drug candidates interact with human biological systems. The resulting tools are intended to support regulatory decision-making and improve the efficiency of drug development while enhancing patient safety.
ARPA-H’s funding reflects growing federal interest in modernizing drug development workflows, particularly through the use of artificial intelligence and human-relevant testing systems. The agency has identified alternatives to animal testing as a priority area for accelerating innovation and improving translational success rates.
Amgen stated that participation in the program aligns with its broader strategy to adopt advanced technologies that improve research productivity and reduce development risk. The company has increasingly explored AI-enabled platforms to complement traditional experimental approaches in discovery and development.
If successful, the collaboration could help establish a new framework for preclinical safety testing, supporting regulatory acceptance of AI-based models and human tissue data as credible alternatives to animal toxicology studies.
By