M E E T I N G S
Plenary Session #3: The AI Landscape: Threats, Challenges, and Opportunities
The Commissioners met for the third plenary session on July 11, 2019, in Cupertino, California.
During their day-long meeting, they received presentations from the Intelligence Community, the National Security Council, and the Department of Defense on threats to AI research and development in the United States, and the White House’s progress in developing an action plan to implement the National Security Presidential Memorandum on AI.
The working group chairs provided readouts from their working group meetings. Commissioners reviewed, discussed, and approved the draft Initial Report to Congress due at the end of July, and then discussed the path forward for submitting the Interim Report to Congress in November. They concurred and approved the Commission staff’s recommendation to organize a public conference that mirrored the Commission’s working group structure, making it a total of four panels with opening, lunch, and closing speakers. Each panel will include leaders from academia, industry, and government. The Commissioners agreed that the public conference is a great way to engage a wider audience and solicit feedback on the Commission’s initial assessments. The purpose of the conference is to build upon the ongoing public discourse around AI and focus on the AI-related national security issues as detailed by Congress in the legislation that created the Commission. The Commissioners selected November 5, 2019 for the date to hold the conference. At the end of the day, the Commissioners received a briefing from a subject matter expert on international workforce and talent trends.
As the next couple months would be focused on the Interim Report, the Commissioners agreed to hold their fourth plenary session in Washington, DC in October.
Plenary Session #2: AI, the Global Environment, and U.S. Government Strategies
On May 20, 2019, AI experts from the White House Office of Science and Technology Policy, the National Security Council, and the Defense Department’s Office of Net Assessment briefed the Commissioners on their current strategies and views on AI-related national security issues from foreign policy, security, and defense-related perspectives.
After receiving briefings, the Commissioners met in their working groups with staff to discuss progress on the staff’s research and analysis. During its executive session of all Commissioners, they agreed that there should be three phases to get the Commission to its final report: the assessment phase, where Commissioners and staff will be in listening mode, gathering information from a variety of experts; the analysis phase, where Commissioners and staff develop initial judgements; and the findings phase, where Commissioners and staff refine their assessments into the final recommendations to Congress. These phases will meet the suggested reporting deadlines to Congress--an Initial Report due in July, an Interim Report due in November, and a Final Report due in March 2021.
The working group chairs provided readouts from their working group meetings. After previously organizing themselves into five working groups, the Commissioners agreed to streamline their focus areas into four working groups. They also endorsed the staff’s recommendation to establish three staff-led Special Project Groups that will focus on cross-cutting issues important to all of the working groups: public-private partnerships, ethical development and use of AI, and data for national security.
The Commissioners asked the staff to present a draft of the Initial Report for discussion and approval at the next plenary meeting, and they agreed their next meeting would be in July. Before closing, the Commissioners received a briefing on the AI tech landscape.
Plenary Session #1: Setting the Business Rules, Initial Briefings
Newly-elected Chair Eric Schmidt and Vice Chair Bob Work called to order the first plenary session of the National Security Commission on Artificial Intelligence on March 11, 2019 in Arlington, VA. The Chair and Vice Chair opened the meeting by outlining the objectives of the first plenary session. First, Commissioners needed to agree on an operating plan and responsibilities, and second, agree on an approach to Congress’ detailed scoping of the Commission’s review as it researches, analyzes, and considers legislation, policy and outreach opportunities to advance the goals of the Commission.
The Commissioners received presentations from the Department of Defense, Department of Commerce, the Intelligence Community, Members of Congress, and Congressional staff on their perspectives on AI threats, the National Defense Strategy, the role of Congress, the Defense Department’s AI organization, export controls, and best practices from past commissions. The newly appointed Executive Director and Chief of Staff briefed the Commission on their plans to stand up the office and the work completed to date.
Commissioners voted in favor of holding plenary sessions every other month. Each plenary will be a full day. They approved establishing five working groups structured around broad categories. Each working group plans to meet on alternate months, and will be comprised of a subset of the Commission’s members and staff. The Commissioners self-selected the working groups on which they would serve. Commissioners agreed that the working group categories will be AI research; national security applications for AI; AI for the betterment of citizens and the economy; preparing citizens and the workforce for AI; and ensuring international competitiveness and cooperation in AI.
Commissioners agreed to rotate meeting locations to maximize convenience and to allow the Commission to engage with AI communities around the country. They agreed to hold their next plenary meeting in Cupertino, California on May 20, 2019.