Some academics argue that mind-uploading is completely impossible, with hypotheses such as the fact that a mind can only exist in its original environment or that the upload will only appear as a frozen copy from one specific second of time. And as such, I have to ask whether an incomplete and non-human iteration of a mind or a segment of an individual is capable of being that individual and predicting their behavior.
The core of behavior interpretation relies on understanding how patterns, interests, experiences, and personal traits dictate choices, decisions, and actions. Such strategies are implemented in psychology all the way through current advertising on the internet based on previous searches, private data, and other fragments we leave every time we use our devices. It is currently being used for and against us, but not to perfection. As such, I think computer simulation will be able to generate consistent and generally accurate predictions; after all, it is in possession of vital information. Perhaps we are not even conscious of some of the data it will receive.
However, the predictions will be incomplete, either due to the lack of usual external factors or the imperfect recollection of the past. As such, does such a model allow for the free will of this computer? The free will of us within the universe cannot be stated with certainty, but free will in relation to us understanding our own consciousness is possible. As we are conscious, we often believe that it allows for freedom of choice, and as such, if the computer emulation claimed to be conscious, would it make it so? It may have the capabilities to make this claim, but is there any way to test it? I think that our understanding of free will is still very hypothetical, as there is the possibility that every aspect of our lives is driven by norms taught to us, deities, butterfly theory, or something completely unknown to us still. The chance of complete free will is just as likely.