Remarks by Commerce Secretary Gina Raimondo at the AI Safety Summit 2023 in Bletchley, England

Remarks by Commerce Secretary Gina Raimondo at the AI Safety Summit 2023 in Bletchley, England
KCPullen@doc.gov
Thu, 11/02/2023 – 13:10

Cybersecurity

AS PREPARED FOR DELIVERY

Thursday, November 2, 2023

Office of Public Affairs

publicaffairs@doc.gov

Gina M. Raimondo

Thank you to Secretary Donelan and Prime Minister Sunak for bringing us all together. It is inspiring to see so many different sectors and countries represented to discuss one of the most important challenges facing us all.

Bletchley Park is a symbol of technology’s potential to empower the heroic side of humanity and how advancements in technology can change the course of history. It is also a reminder that women – who were three-quarters of the Bletchley Park codebreakers – have played an essential but underappreciated role in advancing science and harnessing its benefits. Today we are here to address the unique dangers of the most advanced AI, due both to misuse and risks these systems may not operate as we intend. The Biden Administration’s approach is to manage AI’s risks so that we can maximize its potential.

This approach is reflected in the voluntary commitments secured by the White House, under which U.S. AI companies committed to safe, secure, and trustworthy AI development, including measures to rigorously test, or ‘red-team’, AI models, mitigate harmful capabilities, and protect the AI models from unauthorized release.

President Biden’s new Executive Order on AI, published earlier this week, builds on these commitments by directing the Department of Commerce to ask frontier AI developers to disclose red-teaming, safety, and cybersecurity measures taken for next-generation frontier models. We see this combination of voluntary measures coupled with reporting requirements as an important step toward the safe development of frontier AI.

Well-developed standards and testing capability will also be critical for safety. I welcome the UK’s launch of its AI Safety Institute; and am pleased to announce that the U.S. will create its own AI Safety Institute, to be housed at NIST within the Department of Commerce, to fulfill priorities of the Executive Order. The NIST-led Institute will facilitate the development of standards for safety, security, and testing, and provide testing environments to evaluate both known risks and emerging risks at the frontier. I’m also glad to announce that the Institute is launching a consortium to work with partners in academia, industry, and non-profits to advance its frontier AI safety mission. We intend to build on NIST’s long history of producing scientifically grounded definitions, metrics, and frameworks, which have informed domestic and international progress, for the common good. We also commit to establishing a formal partnership between our Safety Institute and the United Kingdom’s. We are eager to begin our work together.

But this work does not end with the U.S. and the UK. We are working to expand information-sharing, research collaboration, and ultimately policy alignment across the globe on advanced AI. 

Even as nations compete vigorously, we can and must search for global solutions to global problems. In the nuclear age, superpowers came together to cooperate and formulate global mechanisms of assurance. In the age of AI, we have an even more complex problem that involves all nations and the private sector. This will require global coordination on AI safety. It will require the commitment of every country to prevent misuse and ensure that dangerous AI technologies do not fall into the wrong hands. We cannot be daunted by that. Instead, we must be called to action, together. Thank you.

Bureaus and Offices

National Institute of Standards and Technology

Leadership

Gina M. Raimondo

Tags

Artificial Intelligence

Read the full report from the U.S. Department of Commerce: Read More