But mostly, I just think “my AI says its trending toward significance” belongs in the bad stats take hall of fame.

Not sure whether German data allows such an analysis 🫠

For mothers this is possible for a number of countries (see long-term trends for France). For fathers, only few statistical offices collect information on paternal age at childbirth by parity. The longest of these time-series only go back to the early 1960s:
Norway: www.ssb.no/en/statbank/...