November 17, 2023
Remarks as prepared
Good morning, everyone! Thank you all for joining us here at the Commerce Building for today’s U.S. AI Safety Institute workshop. You all represent the future of AI — its capabilities, its evolution, and its potential.
This convening will serve to build upon a long tradition of collaboration in creating trustworthy AI. Your help with the AI Risk Management Framework is an essential component of its success, and your engagement with our Generative AI Public Working group has showcased just how committed you are to this critical issue.
But you’re not alone. President Biden’s historic signing of the AI Executive Order late last month has helped launched us on the path toward an AI ecosystem that will manage the risks so we can harness the benefits. And the effort that brings us all together today is central to the President’s vision.
Our approach to realizing that vision is straightforward and comprehensive: to combine ambitious voluntary measures — including domestic and international voluntary commitments — with sophisticated standards and evaluation capabilities.
That’s why Secretary Raimondo attended the APEC Leader’s Meeting in San Francisco this week, where she met with foreign officials from across Asia about our desire to move together in lock step towards the goal of global AI policy alignment. And before that, the Secretary traveled to London for the UK AI Safety Summit, where she highlighted the Executive Order, as well as the adoption of the G7 Principles and Code of Conduct on AI.
What’s more, earlier this month Vice President Harris and Secretary Raimondo announced a new resource on AI that we will seek to leverage: the US AI Safety Institute. The Institute will:
And that’s what brings us here today. This convening exists because the Administration understands just how much these discussions on AI matter. We can and we must work together in this US AI Safety Institute Consortium. Our consortium’s work will not only be open and transparent, but it will provide a hub for partnership in building and maturing a new measurement science for trustworthy AI.
Through this Consortium, we will develop new guidelines, tools, methods, protocols and best practices to facilitate the evolution of industry standards for developing or deploying AI in safe, secure, and trustworthy ways. From guidance around AI capabilities, authentication, and workforce skills, to generative AI approaches and ensuring the availability of adequate testing environments, we will seize the promise of AI, manage the risks and build a safer, more secure world.
Now, what I have just listed can seem like a daunting set of challenges for us to address; however, I know we are capable of seeing it all through — and, together, I know we will.
Our goal here is to put forward the foundation for a long collaboration and the long road ahead. We look forward to facilitating lively discussion of these issues. Ultimately, your thoughts, time, and effort are what will make this a success.
November 17, 2023 remarks during the USAISI Workshop on Collaboration to Enable Safe and Trustworthy AI held at the U.S. Department of Commerce Washington, D.C. and online.