Can the risks associated with Shadow AI be turned into opportunities for responsible technological advancement?