Human rights alignment: The challenge ahead for AI lawmakers

Abstract

The frameworks for the governance of AI have evolved rapidly. From the 2018 Universal Guidelines for AI on through the 2019 OECD/G20 AI Principles 2019, and the 2021 UNESCO Recommendation on AI Ethics, governments have agreed to the basic norms to regulate AI services. Two important legal frameworks are also now underway-the EU AI Act and the Council of Europe AI Convention. As these frameworks have evolved, we see the scope of AI governance models expand. From an initial focus on "human-centric and trustworthy AI" through the recognition of "fairness, accuracy, and transparency" as building blocks for AI governance, we see now consideration of sustainability, gender equality, and employment as key categories for AI policy. AI laws also overlap with familiar legal topics such as consumer protection, copyright, national security, and privacy. Throughout this evolution, we should consider whether the evolving models for the governance of AI are aligned with the legal norms that undergird democratic societies-fundamental rights, democratic institutions, and the rule of law. For democracies to flourish in the age of artificial intelligence, this is the ultimate alignment challenge for AI.