Faculty Director Gary Marchant recently wrote on approaches to the governance of artificial intelligence (AI) for UCLA School of Law’s new AI Pulse project. Marchant’s article, “Soft Law” Governance of Artificial Intelligence, takes a look at calls for traditional governmental regulation of AI, and where such an approach is unsuitable for AI technology and development.
Marchant instead proposes that Soft Law approaches have advantages over traditional governmental regulation and have been growing in popularity and application.
“Soft law are instruments that set substantive expectations that are not directly enforceable by government. They can include private standards, voluntary programs, professional guidelines, codes of conduct, best practices, principles, public-private partnerships and certification programs. Soft law can even include what Wendell Wallach and I refer to as “process soft law” approaches such as coding machine ethics into AI systems or creating oversight systems within a corporate Board of Directors.”
Marchant describes current soft law and AI efforts from the Institute of Electric and Electronic Engineers (IEEE), the Partnership on AI, and the Future of Life Institute, as well as efforts by specific companies. He then evaluates the strengths and weaknesses of existing approaches and proposes the creation of a entity tasked with coordinating the work of different soft law organizations focused on AI – a Governance Coordinating Committee.