It's my understanding that ChatGPT isn't very good with facts and figures.
ChatGPT, at least in its current form, isn't supposed to do anything more than being able to sustain a "conversation in natural language".
It's not compelled to fact check, to be creative and original, to debunk inaccuracies from the person it's talking to.
Nothing of what it says can be taken at face value as "fact-checked" and you can easily prompt the AI to contradict itself in the span of few lines.
For instance if you "scold" ChatGPT for saying something you don't like, it will apologize and say it was a mistake even when it just stated something factually correct.
