I/ITSEC NEWS: Data Chiefs Sound Alarm on Chinese AI Ambitions
ORLANDO, Fla. — Uniformed and civilian military officials are worried about the U.S. government’s ability to defend against China’s artificial intelligence-enabled capabilities.
The Defense Department could lose its competitive edge over China if policies remain the same and progress is not accelerated, officials said Nov. 30 during a panel discussion at the National Training and Simulation Association's annual Interservice/Industry Training, Simulation and Education Conference in Orlando, Florida. NTSA is an affiliate of the National Defense Industrial Association.
A recent independent report by the National Security Commission on Artificial Intelligence called for speedy upgrades to the Pentagon's data architecture. The commission was established under the fiscal year 2019 National Defense Authorization Act to examine ways to advance the development of AI for national security and defense purposes.
The nearly 800-page document found that the U.S. government is not currently prepared for AI threats, and must become “AI-Ready” by 2025.
“It's an unclassified report, so just imagine what the classified report says,” said Sae Schatz, director of the Advanced Distributed Learning Initiative, which advises the Pentagon on emerging technologies.
Schatz pointed to the report’s assessment that if China outpaces the United States in producing the best algorithms, the United States and its allies will face a “geopolitical challenge” from Beijing.
“We all need to get on with it,” she said, referring to the 2025 "AI-Ready" deadline.
Meanwhile, Deputy Secretary of Defense Kathleen Hicks signed a memorandum in May promising to transform the Defense Department into a “data-centric organization” by 2025.
Col. Ryan Kehoe, deputy chief data officer for the Air Force, said he is worried about meeting the accelerating timeline for developing advanced capabilities.
“I've heard dates [that say] 2035. Then I heard 2028, and now I hear 2025,” Kehoe said. “I'm looking at my watch — it's almost 2022 — how are we going to get there in that timeline?”
The Defense Department's funding processes makes it challenging to develop more advanced data architectures, he noted. For example, the continuing resolution that the federal government is currently operating under prevents new programs from starting across the Pentagon.
“Right now, we're sitting in a [continuing resolution] and we can't act on the things that we already planned to do," he said. "It's very frustrating."
China’s use of artificial intelligence to operate unmanned tanks and surface vessels as well as its significant investment in research "should keep us up at night,” said Portia Crowe, chief data strategist at Accenture Federal Services and former chief data officer for Army Futures Command.
“We're making great advances, but I think that there is a lot more room to grow together by 2025,” she said. “We've got to move a little bit faster.”
Tom Sasala, chief data officer at the Department of the Navy, said experimentation and exploration with some risk is necessary if the Pentagon wants to beat China’s tech acceleration. He noted that some military leaders may be reluctant to share their data more broadly within the Defense Department because of the fear of failure.
China’s access to huge volumes of data is already setting the stage for the nation to take on the United States, he noted.
“Someone mentioned earlier ... the amount of overmatch that we have right now at the moment,” he said. “It's not a lot, and in some cases, it is not good insofar as we are outmatched.”