As artificial intelligence (AI) becomes more integrated into intelligence operations, U.S. agency officials face a challenging balancing act. While AI promises improvements in data analysis and automation, use of the technology also poses significant concerns about privacy, civil liberties, and potential misuse by adversaries.
Intel officials from several agencies hashed through those challenges during a webinar hosted by the Intelligence and National Security Alliance (INSA) on Aug. 20, where several highlighted the potential of AI to improve data analysis and speed up pattern recognition.
At the State Department, for example, AI has become a crucial tool in gaining insight into the mass quantities of data at hand. According to Lisa Kenna, principal deputy assistant secretary intelligence and research at the agency, AI technology helps make sense of all that data.
“The intelligence community should leverage cleared AI tools to accelerate analysis, data organization, and summarization. Many analysts are already using these tools and seeing significant potential for their broader application,” she said.
“This is one of the reasons why we created an open-source coordination unit that is set up to help our analytical teams think about how to engage with other [intelligence community] partners in industry to make most efficient use of the data,” Kenna said.
Other panelists stressed the importance of addressing privacy concerns and upholding civil liberties in the decision to leverage AI tech.
“To some degree, we’ve already got some aspects of AI already out there, but understanding where it’s going to take us is really the hard problem,” said Jay Tilden, director of the Office of Intelligence and Counterintelligence at the Department of Energy.
According to Tilden, the intelligence community needs a clear strategy for using AI in compliance with all requirements while safeguarding the privacy and civil liberties of citizens.
“How can we ensure that AI uses the correct data sets to improve analysis and automate redundant tasks? At the same time, we must consider the potential misuse of AI by adversaries,” Tilden said. He added many of the current administration’s policies are aiming at achieving both of those goals.
Kenneth Wainstein, the undersecretary for Intelligence and Analysis at the Department of Homeland Security, echoed Tilden’s message by adding the “AI is both an opportunity but also a potential peril.”
“We have stood up an AI task force at the department to look at how AI can be used effectively, whether it’s looking for trends of law or patterns of malign conduct. AI is there, especially in the intelligence space. [But] you must be very careful about using it in a way that protects privacy. So, we’re busily working on our policies about how it can and should not be used by us,” Wainstein said.
