AI & Analytics
Pentagon needs AI on each leadership level, panel says
- By Lauren C. Williams
- Mar 16, 2021
The Defense Department needs to embrace artificial intelligence at every leadership level, according to Robert Work, vice chairman of the National Security Commission on Artificial Intelligence and former deputy defense secretary.
"We are shifting into an era of systems warfare," Work testified during a joint hearing with the House Armed Services Subcommittee on cyber, innovative technologies and information systems and the House Oversight and Reform national security subcommittee on March 12. "Our adversaries explicitly say this and say the way we will defeat the U.S. military is to have better operational systems and the way to get there is to inject AI applications and autonomy into the systems."
The commissioners testified, as recommended in its nearly 700-page report, that DOD should intertwine AI throughout its governance structure for the U.S. to gain technological advantage over adversaries.
The commission recommended more funding for the Defense Advanced Research Projects Agency to "coordinate multiple research programs to detect, attribute, and disrupt AI-enabled malign information campaigns and to authenticate the provenance of digital media" and transition the tech and applications to government departments and agencies.
"This will literally affect every operation, mission that we do. And it's going to require a different way of training our and educating our commanders and our people," Work said.
The former deputy defense secretary also highlighted the need for a common digital infrastructure, such as the Joint AI Center's Joint Common Foundation, with the private sector, education and research institutions through a secure cloud that would house algorithmic libraries and using cross-functional teams to look at systems components, from sensors to command and control, to figure out how AI can help decision making.
Eric Schmidt, the NSCAI chair and former Google CEO, said DOD will have to make a "much stronger" statement than the JAIC to adopt AI on a broad scale, and that "having AI be a major component of strategy at every level is one of the ways that the top-down [approach] can work."
But there also needs to be an approach from the bottom that includes a strong statement on AI from the Pentagon, he said.
"I would like to see a DOD statement around AI that is much stronger than we currently have with the JAIC," Schmidt said, including "a higher reporting level, more resources and...individual control at the commander level and the [combatant command] level where they have flexible teams which can be used to solve important national security problems [and] will include AI expertise."
Chad Cisco, the general manager of Federal business at DataRobot told FCW that such a shift will require a pivot in how resources are spent.
DataRobot explored where DOD's AI money was being spent, and found after looking at about 600 projects and $6 billion allocated for fiscal years 2020 and 2021, the vast majority of spending went to early stage AI development tools with less than 1% for AI system development or operating systems, Cisco said.
"Leaders at the top level have to understand what is a good application of AI" Cisco said. "They can't contract this out to systems integrators if they really want to make the change they need."
This article first appeared on FCW, a Defense Systems partner site.
Lauren C. Williams is senior editor for FCW and Defense Systems, covering defense and cybersecurity.
Prior to joining FCW, Williams was the tech reporter for ThinkProgress, where she covered everything from internet culture to national security issues. In past positions, Williams covered health care, politics and crime for various publications, including The Seattle Times.
Williams graduated with a master's in journalism from the University of Maryland, College Park and a bachelor's in dietetics from the University of Delaware. She can be contacted at [email protected], or follow her on Twitter @lalaurenista.
Click here for previous articles by Wiliams.