Four Years After Congress Dropped the Animal-Testing Requirement for New Drugs, the FDA Finally Updated Its Rulebook to Match

Here is a number the drug industry has lived with for two decades. Of the drugs that pass preclinical animal safety testing and enter the first phase of human trials, only about 6 to 10 percent ever reach approval. When they fail, they fail on the two things the animal was supposed to predict: across a large sample of early trials, lack of efficacy drove about 60 percent of terminations and toxicity another 30 percent. A mouse is not a small human, and a beagle is not a large one. The animal model has always been a proxy, and a leaky one, and the leak has been showing up in the attrition data for years.

So when the FDA announced on September 21 that it had issued a direct final rule to “advance innovative alternatives to animal testing”, the honest question was never whether the old paradigm deserved to go. It did. The question is what this particular action actually does, and whether the headline is running out ahead of the fine print. Read the rule and the answer is smaller than the press release: the FDA changed the words, not the bar.

What it primarily does is edit vocabulary. Across the regulations, “animal tests” and “animal studies” become “nonclinical tests” and “nonclinical studies,” and the language implying that animal testing is the only acceptable way to generate pre-human safety data comes out. The methods it blesses by name are the ones you have heard about: human cells, organs-on-chips, computer models, the family of techniques the agency files under New Approach Methodologies. Acting Commissioner Kyle Diamantas framed it as supporting “the Trump Administration’s push” to complement or, where appropriate, replace animal studies.

Then comes the sentence the coverage tends to skip. By the agency’s own account, the rule “does not eliminate or prohibit animal studies, change evidentiary standards or impose new costs or requirements on drug developers.” In plain terms, nobody has to do anything differently tomorrow than they did last week. A sponsor who wanted to run an organ-chip study instead of a rodent one could already make that case. A sponsor who wants to keep dosing beagles still can. What changed is the wording in the rulebook, not the evidence a drug has to bring.

And it changed late. The real legal turn came almost four years ago, when the FDA Modernization Act 2.0 became law at the end of December 2022, stripping the decades-old statutory requirement that new drugs be tested in animals and explicitly authorizing cell-based assays and computer models as alternatives. Congress did the heavy lifting. This week’s rule is the agency finally editing its own regulatory text so it stops contradicting a statute that has been on the books since the Biden administration. That is not nothing, because regulations that lag the law leave reviewers and applicants guessing. But it is bookkeeping, and it is worth naming as bookkeeping rather than a breakthrough.

The breakthrough, if it is coming, was mapped out separately and earlier. In April 2025, then-Commissioner Marty Makary released a three-year roadmap to make animal studies “the exception rather than the norm”, starting with monoclonal antibodies, where he argued animal testing “has a poor track record of predicting safety and efficacy in humans.” He is right about the track record. The open question is whether the replacement is ready, and here the instinct to distrust official confidence points exactly where the boosters would rather you not look.

Because the alternatives are not validated either. Around the one-year mark of the push, outside experts cautioned there was still a long way to go: the field still lacks the shared benchmarks and formal validation studies that would let a regulator know when one of these platforms is actually predictive rather than merely impressive. The considered read from the people building the tools is that these methods work best as supplements to established testing for now, not drop-in replacements. That is the sober version the announcement’s momentum tends to flatten.

So watch where the benefit lands. Retiring a wasteful animal-testing mandate is a win, the kind of deregulation that serves patients rather than incumbents. But swapping an old proxy everyone knew was flawed for a new proxy nobody has yet proven, and calling the swap “innovation,” is how a fresh layer of unvalidated gatekeeping gets sold as progress. The organ-chip makers and the in-silico and AI-modeling shops have an obvious commercial interest in “nonclinical” becoming the default word, and that interest is worth keeping in frame. To showcase the case, the FDA also launched a database of 25 NAM use cases drawn from its own past review materials. That is a curated argument, not an independent validation record.

The next real signal is not this rule. Because it was issued as a direct final rule alongside a companion proposed rule, the agency has committed to withdrawing it and reopening the standard notice-and-comment process if it receives significant adverse comment, so the comment file over the next few weeks is the first thing to watch. The second is the monoclonal-antibody phase of Makary’s roadmap. If that produces published, benchmarked methods that predict human outcomes better than the mouse did, the terminology change will have earned its billing. Until Diamantas and the agency put that evidence on the table, the plain reading holds: the words changed this week, and whether the science changes with them is a question the next round of data will answer, not this rule.

Sources

1. FDA – Updates Regulations to Advance Innovative Alternatives to Animal Testing (Sept 21, 2026)
2. FDA – Plan to Phase Out Animal Testing Requirement for Monoclonal Antibodies and Other Drugs (April 2025 roadmap)
3. Fierce Biotech – FDA celebrates progress to end animal testing but experts warn there’s still a long way to go
4. PMC – FDA Modernization Act 2.0: transitioning beyond animal models with human cells, organoids, and AI/ML approaches (preclinical-to-approval attrition data)
5. PubMed – FDA Modernization Act 2.0 allows for alternatives to animal testing (Dec 2022 law)
6. FDA – New Approach Methodologies (NAMs) resource and use-case database