I'm kind of a dinosaur, and I didn't know that 'AI interviews' exist. But if they are conducted via Zoom, Google Meet, or Microsoft Teams then yes, the FDT would work. No modification is needed at the other end of the video conference system when using the FDT. In fact, such AI-based interactions might well obscure a small limitation of the FDT, namely that there is 'latency', or temporal gaps, between your speech and the broadcast of your reconstructed, fluent speech to Zoom. The latency is only a few seconds, which people notice in a big way, but maybe not an AI 'interviewer'.