A group of leaders and scholars from MIT has released a set of policy briefs aimed at developing a framework for the governance of artificial intelligence (AI) in the United States. The goal of this framework is to enhance US leadership in AI while mitigating potential risks and exploring the benefits of AI deployment.
The main policy brief, “A Framework for U.S. AI Governance: Creating a Safe and Thriving AI Sector,” suggests AI tools can be overseen by current governing bodies and recommends clearly identifying the purpose of AI devices for regulation to be appropriately implemented. The researchers encourage initializing with areas where human activity is already regulated as they pose high risk.
The policies come during an increased interest in AI, with substantial investments into the sector. The European Union is also working on finalizing their own AI regulations, highlighting the challenge of governing both general and specific AI tools.
The MIT team was keen to get involved, given its expertise in AI research. It emphasizes the need for AI providers to define intent and purpose for AI applications in advance. This would allow for existing sets of regulations and regulators to be relevant to each AI device. The guidelines outline how current policies could adapt to include AI, using existing regulatory agencies and legal liability frameworks.
The report discusses potential problems when applications exist on multiple levels or “stacks”. It suggests that providers of specific services might primarily be liable for issues. However, the makers of general-purpose tools should also hold some accountability if their technologies contribute to problems.
The policy brief calls for advances in AI tool auditing, with established public standards. It also suggests the creation of a new government-approved “self-regulatory organization” to accumulate domain-specific knowledge and respond flexibly to the dynamic AI industry.
The policy papers explore a range of regulatory issues and encourage further research into how AI can benefit society. This includes the examination of how AI might augment and aid workers, rather than replace them, promoting inclusive economic growth.
The committee aims to balance the enthusiasm and concerns about AI by advocating for appropriate regulation to accompany technological advances. They emphasize the important role educational institutions have to play in offering expertise about technology’s interplay with society. The intention behind the work is to ensure that AI serves the nation and the world effectively and responsibly.
