A commission assessing the best path forward for the U.S. government and national security community in the area of artificial intelligence (AI) on Monday unanimously approved its latest set of recommendations as it continues its work toward a final report next March.

The 35 recommendations approved by the National Security Commission on Artificial Intelligence (NSCAI) fall under six lines of effort, including research and development and software, application to national security missions, training and recruiting talent, protecting and building on U.S. technology advantages and hardware, marshaling global cooperation, and ethics.

A draft report released in conjunction with the NSCAI’s July 20 meeting says the Defense Department needs to reform its research and development “ecosystem to enable the speed and agility needed to harness the potential of AI and other emerging technologies,” including making it easier to share software, components and infrastructure and the creation of an AI software repository.

The draft report also calls for improvements in how DoD tests and evaluates AI-based technologies.

Another recommendation calls for creating a Bureau of Cyberspace Security and Emerging Technology within the State Department, a move that would consolidate existing, but disparate functions, allowing it to better compete for resources and “serve as a focal point and champion for security challenges associated with emerging technologies,” it says. The recommendation is based on a department proposal.

The draft also recommends a senior leadership council withing the State Department “To strengthen the link between great power competition strategy, organization, foreign policy planning, and AI.”

The NSCAI was established in the fiscal year 2018 defense authorization act.