How do we test for intelligence? Well, suppose that we make an observation of an ordered repeatable event. Since the event is ordered it is not random and since it is not random there exists a causal pathway that it takes to produce the observed result.
We can study this pathway. Since we are able to make some observations about the phenomenon we can safely conclude that we have some measure of intelligence. If we are unable to completely unravel the causal pathway despite our best efforts then we can conclude that the measure of intelligence that assembled the causal pathway surpasses our intelligence and evidently some intelligence was employed in the genesis of the observed phenomenon.