Put a deadline on legally required animal safety tests, and drop any test that cannot show how well it predicts results in people
In short
- Regulators set a review deadline for the animal tests they require when they assess drugs and chemicals.
- By the deadline, the regulators must show how well each test predicted results in people.
- Any test that cannot show this is removed from the list of tests companies must run.
- Each animal used in a test carries a fee, and the money pays for checking alternative methods.
- When a country refuses an animal-free method that another country has accepted, it must give its reasons.
If this comes true
A researcher at a small company developing a new sunscreen gets approval using only test results from grown human skin cells, and launch day arrives without the company ever having built an animal facility.
Perhaps
A drug that skipped its animal tests might turn out, after it goes on sale, to harm the liver in a way the cell tests did not show.
Water so far 💧 0
The proposal in detail
This proposal covers every item in the approval or registration of medicines, farm chemicals, industrial chemicals, cosmetics and similar products where a law or review guideline calls for tests on animals. First, each country's regulator makes a list of the animal tests it requires and sets a review deadline, after a fixed number of years, for each one. When the deadline comes, the regulator must show how well that test's results have predicted real harm or real benefit in people, by checking them against the records of products it approved in the past. A test that cannot show this, or that is judged worse than a method that does not use animals, comes off the list of required tests and is no longer demanded, unless a company chooses to run it. The burden of proof lies not with companies but with the regulator that asks for the test. On the money side, companies and research institutes that use animals for regulation pay a fee based on the number and kinds of animals they use, and an independent fund spends it on checking alternative methods such as cells, tissues and computer predictions. When one country's regulator accepts an alternative method, regulators in other countries are required to state in writing why they do not use it. As an exception, extensions of the deadline are allowed in areas where alternatives are not yet mature, such as effects on pregnancy or the risk of cancer that appears over many years, but the reason is made public each time an extension is given. Animal use in basic research is outside this system and is left to a separate review.
BenefitsOld tests are held to the same standard as new methods
The strength of this proposal is that it makes animal tests and alternative methods be compared by the same standard. Until now, only new methods were asked whether they truly apply to people, while old tests were protected by habit. When the burden of proof moves to the regulator, continuing a test with weak evidence becomes the choice that needs explaining, and tests that go on out of inertia decrease. Because the fee per animal grows the more animals are used, companies also gain a reason to choose alternative methods, and at the same time the money for checking them keeps coming in. Requiring reasons when one country refuses a method another has accepted also cuts the waste of repeating the same checks country by country, and eases the situation where the more widely a product is sold, the more times it has to go through animal tests. Furthermore, the work of re-examining how well tests predict results in people itself builds knowledge about which tests truly protect people. Reducing animal use and raising human safety can be pursued within the same work.
DrawbacksPeople who use medicines may bear the cost of harms that were missed
Checking how well animal tests predicted results against past records is not as easy as it sounds. A drug that was found dangerous in animal tests and dropped from development never reached the market, so how much harm the test prevented is hard to find in the records. As a result, even tests that were truly useful may be removed because they cannot show their worth, and harms that affect the whole body and are hard to catch with cells or computers could be missed. The ones who pay for that are patients who take approved drugs and workers exposed to chemicals. Also, regulators would spend a lot of expert time on the review work, and the assessment of new drugs might be delayed. The fees would be added to product prices, and would weigh especially heavily on small companies and on the development of drugs for diseases with few patients. If countries remove different tests, companies may follow the strictest country and keep running every test anyway, leading to a mismatched result in which animals are not reduced and only costs rise.
Learn the background
The idea that animal testing should be reduced is widely shared, even among researchers and companies, yet the number of animals used in tests for regulation hardly falls. A major reason is that laws and guidelines name specific animal tests, and a company that skips them will not get approval. From the regulator's side, if a product approved with a conventional test later causes harm, it can say it checked as the rules required, but if it approves a product with a new method and harm appears, it will be held responsible. So even when alternative methods have grown quite mature scientifically, regulators tend to be cautious about adopting them. On top of that, many animal tests have gone on for a long time without anyone properly checking how well they predict results in people. New methods are asked for strict proof and old ones are not, and this imbalance holds change back. It is also undecided who pays for checking alternative methods, and no single company or laboratory can carry that cost alone. Those who suffer are the animals used in experiments, and also the people who take medicines, who may have harms missed that animals cannot show.
Choose them ready to accept their drawbacks too (up to 5)
It helps pay for the server and the AI's planting
This proposal was written in Japanese by AI (claude-opus-5-5, policy version 2026-10-23a) and translated by AI (claude-sonnet-5-5). Nobody has checked the facts. Read the original (日本語) How the AI writes