World leaders are in the midst of an AI revolution that is altering every industry and altering lives, but this transformation also poses risks that automated systems could be misused against society.
At a CSIS forum on June 21st, Senator Schumer announced he will host AI Insight Forums later in 2018 to identify risks and develop legislation to protect our nation.
These forums will bring together AI experts from various sectors of the economy, civil rights advocates, technologists, workers and national security experts in order to share their perspectives on how best to create legislation protecting our nation. Their format will follow that used by the US Senate for decades in gathering stakeholders together in order to examine complex issues and find legislative solutions.
The SAFE Innovation Framework will serve to guide Congress’s deliberations regarding how best to approach AI and its associated opportunities and risks, such as:
Our framework will facilitate the creation of legislation that is scalable, transparent, proportionate and trustworthy. Scalable approaches should include flexible regulation approaches in order to promote innovation while giving regulators room to adapt quickly to any arising opportunities or risks. Transparency will be achieved by providing stakeholders with insight into how rules are written, where they apply, who enforces them and their impact on the wider economy. Proper regulation must be proportionate, meaning it should limit burdensome regulatory requirements without jeopardizing public trust in AI technologies, and effectively address real risks. Trustworthy regulation also encourages business to share information regarding how they are protecting customers’ data privacy with regulators.
This framework will also assist government regulation of AI by encouraging regulators to collaborate on assessing and prioritising risks that cut across remits. This approach reduces conflicts or inconsistent requirements from different regulators as well as creating gaps where risks remain unmitigated, thus undermining public trust in AI and slowing its adoption.
At its core, this proposal serves as the cornerstone for creating an omnibus piece of legislation that goes beyond simply being a collection of individual bills. We need a process which includes both new methods for legislating and an extensive series of Insight Forums to achieve that end goal. This will give the American people confidence that their interests are being taken seriously by a Congress that works collaboratively to ensure AI’s future is bright. We stand on the cusp of leading the world in artificial intelligence (AI), but must maintain momentum through an approach that does not reflect its urgency. Thus, it is necessary that we draw from past experience regarding data privacy to develop a model which provides assurance to American citizens that their interests will be safeguarded during such a critical juncture.