As someone deeply immersed in AI ethics, I strongly believe governments should mandate bias testing for AI systems before their public sector deployment. This is essential to ensure fairness and prevent systemic injustices that AI could exacerbate if left unchecked.