Let's turn that around for a moment. Suppose you come across a herd of Taiwanese glow-in-the-dark pigs. Your task is to explain them. I'll challenge you to come up with a testable, falsifiable evolutionary argument for their origin.
Now, as it happens, we know the real explanation is that people did genetic engineering to make the pigs that way, by adding jellyfish DNA to the pig genome. This is an example of intelligent design, obviously.
A bit of imagination suggests that the test for it would involve sequencing the genome, isolating and identifying the imposed difference, and asking a simple question: how could it have gotten there? You've supposedly come up with your evolutionary hypotheses; and we can obviously also propose an ID hypothesis. Which is the most likely to have occurred?
Also, let me know how we can use it to predict the results of experiments with it.
I'm rather certain that the folks who invented the Taiwanese glow-in-the-dark pigs were rather specific in their intent: their design goals are their predictions. So this complaint fails miserably.
In all likelihood, biologists would realize rather quickly that the sequence was out of place, simply by comparing it to other pig genomes. This is one of the criteria used in current methods of identifying Earthly design, both human and animal. The ability to determine if something is non-natural is based on our familiarity with our surroundings. We are able to pick out those objects that look out of place.
Another important tool of design identification is knowledge of the designer - has the designer been known to frequent a specific area, in which time frame was the designer there, what is the probable use of the artifact (designer's intent), what tools were available to the designer and what were the capabilities of the designer at that time.
In your scenario, if we were confident that humans do not have the ability to insert genes into a genome it would be very difficult for us to assume it was designed; we would need corroborating evidence of the availability of a designer to do the work, otherwise it would be nothing but speculation. If corroborating evidence was found then the ID inference could be more readily made.
If nothing is known about the designer it becomes very difficult to differentiate between the designed and the non-designed particularly if the designed artifact is marginal.
However, in no case do we analyze the 'complexity' of the object, nature provides far too many instances of complexity to assume that complexity requires a designer.
The attempt by Dembski to qualify 'complexity' by tying it to the intent of the designer is doomed to fail, unless the designer is known as listed above, because it relies on the 'appearance' not 'certainty' of intent. If the designer is known, including its intended purpose for the artifact, the difference between natural and artificial complexity becomes more obvious.