Mitigating Bias in AI Legal Systems: A Framework for Ensuring Fair and Impartial Advice