How can we build AI systems the world can trust? AI2050 Early Career Fellow Nicolas Papernot explores how cryptographic audits and verifiability can make machine learning more transparent, accountable, and aligned with societal values. Read the full perspective: buff.ly/JjTnRjm