That depends on how they're applying their "AI technology." Maybe it's to improve the SNR for a given pixel? Maybe it's to interpolate additional resolution between pixels / lines? It seems possible, but not necessarily likely, that they're doing whole-image post-processing / analysis.