There *are* legitimate use cases for this kind of stuff! Simulation studies! Synthetic data for a preregistered analysis plan prior to an actual study! But this isn't being marketed as that. It's advertised as a cheaper, faster alternative to humans. Temu we-have-results-at-home survey results.
Did you know that from tomorrow, Qualtrics is offering synthetic panels (AI-generated participants)? Follow me down a rabbit hole I'm calling "doing science is tough and I'm so busy, can't we just make up participants?"