AI governance is the idea that there should be a legal framework for ensuring that machine learning (ML) technologies are well researched and developed with the goal of helping humanity navigate the adoption of AI systems fairly. Dealing with issues surrounding the right to be informed and violations that may occur, AI governance aims to close the gap that exists between accountability and ethics in technological advancement. Due to the rise of the implementation of artificial intelligence across sectors everywhere, including healthcare, transportation, economics, business, education and public safety, the concern of definitively outlining AI governance is becoming greater.
The main focus areas of AI governance are AI as it relates to justice, data quality and autonomy. This involves identifying answers to questions surrounding the safety of AI, which sectors are appropriate and inappropriate for AI automation, what legal and institutional structures need to be involved, control and access to personal data, and what role moral and ethical intuitions play when interacting with AI. As a whole, AI governance determines how much of daily life can be shaped by algorithms and who is in control of monitoring it.
Where machine learning algorithms are involved in making decisions, AI governance becomes a necessity. Machine learning biases have been observed to racially profile, unfairly deny individuals for loans, and incorrectly identify basic information about users. The development of AI governance will help determine how best to handle scenarios where AI-based decisions are unjust or contradict human rights.
Organizations that are focused on the future and development of AI governance are the Ethics and Governance of AI Initiative, a joint project of the MIT Media Lab and the Harvard Berkman-Klein Center for Internet and Society, and the White House Future of Artificial Intelligence committee, announced by the Obama Administration in 2016. Both of these initiatives complete significant, public-facing research to investigate and prioritize the societal and political implications of artificial intelligence. However, there still exists a large gap in the legal framework of accountability and integrity of AI.