Artificial intelligence evidence "pushes the envelope" of what courts will accept in order to satisfy themselves that such evidence is reliable and properly authenticated at trial. Courts want to know if artificial intelligence has been used in generating the evidence. They want to know what is in the "box black" and how such evidence was derived and what was it based upon.

Concomitantly, courts need to be comfortable that the results that come from artificial intelligence are reliable and whether the evidence could have been manipulated. These issues beg the question as to whether one's expert can satisfactorily answer these concerns, including whether the use of the particular artificial intelligence-based evidence is generally accepted in the relevant field, before it is permitted to be presented to the jury. That is no easy task! Courts need to be careful in their role as gate keepers, because, once admitted, how a jury could weigh such "provocative" evidence is a slippery slope.