Synthetic tabular data generation addresses data scarcity and privacy constraints in a variety of domains. Tabular Prior-Data Fitted Network (TabPFN), a recent foundation model for tabular data, has been shown capable of generating high-quality synthetic tabular data. However, TabPFN is autoregressive: features are generated sequentially by conditioning on the previous ones, depending on the order in which they appear in the input data. We demonstrate that when the feature order conflicts with causal structure, the model produces spurious correlations that impair its ability to generate syntheResearch goal: How does integrating causal graph structures into synthetic tabular pretraining data affect the zero-shot transfer accuracy of foundation models under covariate shift compared to correlational baselines?Autonomous synthesis report generated by SOVEREIGN Research Kernel. Tribunal consensus score: 8.6/10.