That's not actually true; it causes birth defects in animals as well. Only limited animal testing was performed before it was approved for use in humans; more extensive animal testing, which did demonstrate birth defects, was performed only after it had been withdrawn due to causing human birth defects.
Thalidomide was never approved for treating in nausea in pregnant women in the US. It was prescribed off label, but the FDA rejected it for that purpose.
Why did the FDA reject it if animal tests didn’t show deformities?
To be clear, that’s not a rhetorical question. I’ve heard rumors the FDA wasn’t convinced thalidomide would be effective - which is very different from worries over fetal abnormalities.
Do you know the background there?
I think you inverted the meaning of the parent comment. They asked for exceptions where animal experiments do apply to humans, saying that the rule seems to be they don't.