When talking politics AI is often wrong because it has no idea how often the majority of news outlets intentionally lie. It looks at multiple reports to create it's answers. I could write ten pages about the arguments I've won and forced AI to tell me it was wrong. Supposedly after enough times Grok will remove the outlet from it's list of sources. That said, I did that with CNN and it still uses CNN after telling me if I provide one more example of CNN intentionally lying it would remove them. I did, Grok still uses them. Maybe if enough people did the same AI would start getting more things correct the first time.