U.S. leadership in artificial intelligence will require a whole of government approach that is centered in the White House and includes oversight from the vice president and permanent day-to-day leadership, says the draft final report of a national AI commission.

The Technology Competitiveness Council, chaired by the vice president and managed by a new assistant to the president for Technology Competitiveness, would fill gaps left open between existing White House entities such as the National Security Council, Office of Science and Technology Policy, and others, drive long-term strategy, and “provide a forum for reconciling competing security, economic, and scientific priorities, and elevate technology policy concerns from technical to strategic,” the National Security Commission on Artificial Intelligence, says in the 130-page draft final report.

The commission is meeting on Monday and Tuesday this week to refine the text and make any other adjustments before finalizing the report in March. The commission was created by the fiscal year 2019 National Defense Authorization Act.

The TCC would consist of cabinet secretaries and other White House leaders involved in emerging technologies.

The recommendations on the TCC and vice president leadership were first made in October when the commission published its third quarter report. Another recommendation is for the TCC to develop a National Technology Strategy to guide U.S. policy on all emerging technologies, beginning with AI.

The strategy would build on the major components of the AI commission’s report related to winning the technology competition for guiding U.S. policies, Chris Darby, one of the commissioners and president and CEO of In-Q-Tel, the independent strategic investment firm, said Monday during the virtual meeting. These components, which are included in the draft final report and will be discussed by the commission on Tuesday, are the competition for talent, accelerating AI innovation, intellectual property, microelectronics, technology protection, a favorable international technology order and adjacent technologies.

The draft final report also recommends that the U.S. and China begin a dialogue around science and technology, that is focused on areas of cooperation for emerging technologies where they can “solve global challenges such as climate change and natural disaster relieve,” and also to discuss specific problem areas related to the use of emerging technologies, the draft final report says.

The introduction to the draft report points out the challenges posed by China to “U.S. technological leadership, military superiority, and its broader position in the world.” AI is already being used in disinformation campaigns by China, Russia and others as well as to steal data, it says.

Eric Schmidt, chairman of the commission and former Chairman and CEO of Google [GOOGL], said the introduction in the final report needs to have a “stronger statement” about how close China is to the U.S. in the area of AI. Gilman Louie, one of the commissioners, the founder of In-Q-Tel, and a technology venture capitalist, agreed with Schmidt, pointing out that China is already taking a “whole of country effort” that recognizes AI for its national security, economic and “way of life” impacts.

The commission on Monday discussed the first of two parts of the draft final report, focusing on “Defending America in the AI Era.”

Chapter two of the report is focused on the Defense Department and calls for DoD to have in place by 2025 the foundations to widely integrate AI into its ecosystem, including a “common digital infrastructure that is accessible to internal AI developmental teams and critical industry partners, a technically literate workforce, and modern-AI enabled business practices that improve efficiency.”

Safra Catz, a commissioner and co-CEO of Oracle [ORCL], said that DoD’s current budget, acquisition and development practices for the most part are based on “the industrial age.” She said the Secretary of Defense and Congress will have to be committed to making changes that enable the widespread integration of AI across DoD.

“Bottom-up innovation” also needs to be encouraged, she said.