We'll have bigger worries once AI becomes as good, namely mass unemployment, big tech tyranny etc. For now, you still need to actually reference original sources to be able to verify the (often unreliable) information produced by LLMs.