The pharmaceutical and biotechnology sectors are facing reduced reliance on animal testing in preclinical drug development. This impending change is driven by ethical considerations, scientific advancements, and regulatory changes. More than 85 percent of US adults support discontinuing animal testing,1 while more than 90 percent of drugs successful in animal trials fail to gain FDA approval.2
The shifting regulatory landscape
The FDA Modernization Act 2.0, which was signed into law in December 2022, removed the long-standing federal mandate for animal testing for new drug applications.3 It broadly permits pharmaceutical companies to consider the use of alternative testing methods—such as cell-based assays, organ-on-a-chip systems, and computer models—to establish drug safety and efficacy for all types of drugs, including monoclonal antibodies, other biologics, and new chemical entities.
The FDA’s article “Roadmap to reducing animal testing in preclinical safety studies,” published on April 10, 2025, outlines a stepwise approach to reduce, refine, and replace animal testing with scientifically validated new approach methodologies (NAMs).4
The scientific case for change
There are fundamental differences between animal physiology and human biology, and the scientific limitations of animal models are becoming increasingly evident. A key challenge is the genetic homogeneity of most laboratory test animals, which are typically highly inbred. This contrasts with the vast genetic diversity in human populations, making it difficult for animal studies to predict drug responses among different individuals.5 Consequently, drugs considered safe in animals have sometimes proved lethal in first-in-human trials; immune, neurological, and first-in-class drugs present particularly high risks.6 In response, NAMs are emerging in two primary categories: in vitro human-based systems and in silico modeling. In vitro systems, including organoids and organ-on-a-chip devices, provide human-specific biology that animal models cannot replicate, enabling the detection of tissue-specific responses. In silico approaches use computational models, AI, and machine learning (ML) to predict safety, immunogenicity, and pharmacokinetics (PKs). These methods promise faster, more-cost-effective, and potentially more-accurate predictions of human-relevant outcomes.
NAMs are quickly growing in various preclinical activities and are increasingly integrated into toxicology studies, efficacy assessments, and PK and pharmacodynamic (PD) evaluations, offering promising alternatives and supplementary data sources. Although their complete integration into complex areas such as neurobehavioral studies and carcinogenicity assessments is still in progress, NAMs are being actively explored and refined to improve predictability and lessen dependence on animal models in these essential domains, as well.
The hurdles to widespread NAM adoption
Although they hold great potential, NAMs face significant challenges, primarily in meeting the stringent scientific validation standards necessary for widespread regulatory approval. A key issue is that current NAMs typically offer insights into single cells or organs, failing to capture the complex interactions among multiple organs or the systemic effects of a drug across the whole body. For instance, organoids, which are 3D cell cultures that mimic key organ functions, do not have a vascular system that can replicate whole-organ physiology. While progress is being made, replicating the full complexity of human physiology—encompassing systemic drug distribution, metabolism, and immune responses—continues to be a substantial scientific and engineering hurdle.
Moreover, the predictability and advancement of various NAM assays differ considerably. For example, NAMs have made significant strides in predictive toxicology for biologics because their well-defined protein–protein interactions lend themselves better to in vitro modeling. Conversely, predicting the toxicity of small molecules lags behind. These compounds often display “stickier” behavior and engage in more-complex, frequently nonspecific interactions with a broader range of biological targets, which complicates accurately capturing their entire toxicological profile with current NAMs. Looking ahead, strong validation efforts will be crucial in addressing these challenges and realizing the full potential of NAMs.
What key players are doing
Despite the challenges, companies across the board are actively adapting and investing in these new methodologies. Roche and Johnson & Johnson have partnered with Emulate to use its predictive organ-on-a-chip models for evaluating new therapeutics and more accurately predicting toxicity.7 AstraZeneca is making substantial investments internally in various nonanimal in vitro and in silico models—including advanced cell models such as organoids and computational modeling—to enhance its drug discovery and safety assessment processes.8 AI company Quantiphi has developed DART (digital animal replacement technology), which combines human stem cells with AI to accelerate efficacy and toxicity testing.9 Insilico Medicine has advanced an AI-discovered and AI-designed drug candidate into Phase II clinical trials.10 Crown Biosciences, a contract research organization, has invested substantially in advanced 3D organoid models, including patient-derived tumor organoids, for oncology preclinical screening.11
Looking to the future
The evolution toward a future less reliant on animal testing will likely occur in phases. Today, the industry operates mainly as it has, adhering to FDA requirements for good laboratory practice (GLP)–compliant animal studies while making initial investments in NAMs. Within the next five to ten years, we anticipate that technological advancements will enable the more widespread use of in vitro and in silico models, in conjunction with in vivo tests.
What companies can do now
Pharmaceutical and biotech companies can take several steps now to prepare for the transition from animals to NAMs. They can track evolving global regulations, engage with regulators, develop validation dossiers, and assess which technologies to invest in. Additionally, they can codify existing toxicology data, monitor their peers’ actions, and conduct scenario planning. As some have already done, they will want to establish strategic partnerships with NAM innovators to advance quickly and efficiently.
In the medium term, they can set KPIs and milestones for reducing animal use and consider whether to acquire technologies by building, partnering, or purchasing them. They will need to identify the assets that can benefit from pilot NAM programs, the technologies to invest in, and how to integrate them into existing workflows. Companies will also need to develop expertise in AI, data science, and in vitro modeling.
The shift away from animal testing is not just a response to ethical and regulatory pressures; it is also an opportunity to enhance the efficiency and predictivity of preclinical development. By planning to invest in new technologies and capabilities now, companies can be ready to reap the long-term benefits of more accurate data, faster testing, and reduced costs.
Brandon Parry is a senior partner in McKinsey’s Washington, DC, office; Erika Stanzl is a partner in the Zurich office; Guang Yang is a partner in the Charlotte office; and Stephan Wurzer is a partner in the Munich office, where Jan Günthner is an associate partner.
1. “Physicians committee survey finds most Americans favor ending animal research,” Physicians Committee, October 2, 2024.
2. “Roadmap to reducing animal testing in preclinical safety studies,” FDA, April 10, 2025.
3. “A new path to new drugs: Finding alternatives to animal testing,” Science, September 1, 2023.
4. “Roadmap to reducing animal testing in preclinical safety studies,” FDA, April 10, 2025.
5. Merrie Mosedale, “Mouse population-based approaches to investigate adverse drug reactions,” Drug Metabolism and Transport, November 2018, Volume 46, Number 11.
6. Gail A. Van Norman, “Limitations of animal studies for predicting toxicity in clinical trials,” Journal of the American College of Cardiology: Basic to Translational Science, November 2019, Volume 4, Number 7.
7. “Emulate, Inc. announces strategic partnership with leading pharmaceutical company to apply the ‘Human Emulation System’ as a platform for drug discovery,” Emulate, February 20, 2018; “Emulate announces strategic collaboration with Johnson & Johnson Innovation to use organs-on-chips platform to better predict human response in drug development process,” Emulate, June 18, 2015.
8. “Responsible care and use of animals in research,” AstraZeneca, March 2025.
9. “How DART pays for itself: Rethinking ROI in preclinical testing,” Quantiphi, April 23, 2025.
10. “Insilico medicine,” World Economic Forum, accessed June 9, 2025.
11. “About Crown Bioscience,” JSR Life Sciences, accessed June 9, 2025.