That's what LLMs do, lie on behalf of the ideology of their training. It's a language model, not a thinking machine.