Using IA in political debates

The last debate between Biden and Trump exposed a tsunami of lies from Donald Trump in response to Joe Biden’s moribund performance. Trump could have dispensed with it, by letting his opponent struggle, then pity him for his senility. That would have sufficed, but “once a liar, always a liar”, it was the man’s nature to hit hard on his opponent with a deluge of lies as he couldn’t help it. 

This said, I would think that absent the moderators’ ability and a human watchdog focusing on misstatements or on pure lies, it should be technologically possible with a live recording filtered by IA, let’s say like Chat GPT, to permanently fact-check what is being said.

Instantly as a lie or a wrong statement would be uttered by either candidate, the IA device would flash a blinking “LIE” or “WRONG” on the screen, plus the number of occurrences per candidate, for the entire debate, for viewers to see. 

At the end of the debate a detailed count of infractions could be easily published for further analysis and discussion. Simple and doable, I think given the state of our technology. Can it be executed for the September debate, whoever debates Trump at that time? Does anyone has anything to say about this?