If we use the PC algorithm in the Causal Learn package to do causal discovery with the statement:

"A correlates with B. A correlates with C. A correlates with D. B correlates with C. B correlates with D. C correlates with D. However, B and C are independent given A."

assuming "any relationship not explicitly marked as independent is considered correlated by default". The Markov Equivalence Class (MEC) we get is the following:

If we create a DAG by orienting C->A and A->B, in this DAG, C and D don’t have a confounder. This differs from the result of the generated dataset.