We can always imagine the behavioral/functional phenomena occurring without any corresponding phenomenal consciousness. So this question can never be settled by experiment. But we can develop a theory of consciousness and observe how well the system in question corresponds to the features our theory says correspond with consciousness. Barring any specific theory, we can ask in what ways are the system similar and different from systems we know that are conscious and whether the similarities or differences bear on the credibility of attributing conscious to the system.

Theory is all well and good, but in the end it will have little practical significance. People tend to be quick to attribute intention or minds to inanimate or random occurrences. Eventually the behavior of these systems will be so similar to humans that most people's sentience-attribution machinery will fire and we'll be forced to confront all the moral questions we have been putting off.