For seventeen years, Donald Ingber's lung-on-a-chip — a polymer slab smaller than a USB stick that literally breathes — has been proving that human biology can be modeled without mice. When he first submitted his results to Science, the editors sent him back to run animal experiments anyway. That reflexive deference to animal models is the institutional habit now under sustained assault. The technical portfolio is deep and getting deeper. Organoids mimicking organ structure, organs-on-chips modeling hearts, brains, kidneys, and placentas, multi-organ systems linking up to ten chips to recapitulate human physiology, and AI-driven computational models feeding iterative loops — collectively branded NAMs (new approach methodologies). Emulate, a commercial spinoff, now mass-produces standardized chips with automated culture systems. TissUse in Germany builds multi-organ platforms for pharma clients. These are no longer academic curiosities. The regulatory dam broke in late 2022 when the FDA Modernization Act 2.0 passed into law, explicitly authorizing NAMs for preclinical studies. Previous regulations had mandated animal testing; the new law made alternatives legally permissible. In 2025, the FDA pledged to make animal studies "the exception rather than the norm." By September 2026, the agency issued a rule replacing "animal tests" with the broader term "nonclinical tests" in its drug-development regulations. The NIH now requires grant applicants studying animal models to also incorporate nonanimal research. The shift is global. The European Commission and United Kingdom have announced plans to phase out animal testing. The OECD updated its guidelines to expand NAM use. An estimated 92 percent of all drugs entering U.S. clinical trials fail to reach market — often because animal models failed to predict human outcomes. Failure rates run even higher for heart disease, cancer, and neurological drugs. The economic case for better preclinical tools is enormous. But validation remains the chokepoint. A brain-on-a-chip that works in one lab needs to produce identical results worldwide to earn regulatory trust. Minute variations in hydrogel scaffolds produce wildly different growth patterns. Workflows must be uniform. As Ingber puts it: academic prototypes don't change FDA uptake — commercial products meeting fine performance criteria do. Standardization and mass production are the unsexy prerequisites. TissUse's Ilka Maschmeyer reports a telling inversion: the FDA recently denied a pharma company permission for clinical trials because it presented animal data when the agency wanted organ-on-a-chip data instead. Researchers who spent careers on animal models are now being pushed to add in vitro systems. The regulatory signal is pulling demand, not just permitting supply. Johns Hopkins toxicologist Thomas Hartung frames the real challenge clearly: "It is more about change management than it is about the technology." The technology exists. The regulations are shifting. What remains is the grinding work of standardization, validation, retraining, and cultural transformation across an industry built around animal testing for decades. This is a generational infrastructure transition, not a product launch.
