Charles Lyons-Burt
Government technologists are cultivating a cautiously optimistic view on artificial intelligence, enticed by its many applications while attempting to institute checks and balances so that the tools do not overstep or malfunction.
“AI will revolutionize humanity…but like anything, you don’t go blindly running into it. You run towards, hopefully aware of some of the pitfalls and the challenges,” said Stephen Wallace, chief technology officer at the Defense Information Systems Agency. The Pentagon official contributed to a panel discussion at the Potomac Officers Club’s 5th Annual AI Summit on Thursday.
Wallace said that AI not only relieves users of mundane tasks — it can cut down a 10 minute task to a 30 second blip of responsibility — it also improves accuracy of results and allows users to analyze several systems at once.
Brandy Durham, vice president of the AI practice at ManTech, said it’s fairly well-established that AI is helpful with back office tasks, because there’s less risk involved and the efficiency gains, like Wallace described, are a no-brainer. She also acknowledged that mission use cases are promising for AI implementation. But Durham said she believes digital modernization activities are perhaps the most central use case, both for more “traditional” AI like machine learning and for generative AI.
“A lot of the conversations we’re having are around the automation of human workflows and data workflows, etc., where you can use AI, automation, analytics—whatever the right solution is for that particular use case—to cause an enabling function that maybe our users never see,” Durham offered.
For George Linares, CTO at the Centers for Medicare and Medicaid Services, operational efficiencies relating to privacy and data sharing are a huge net benefit of AI. Linares discussed how, along the lines of his fellow panelists’ remarks, CMS has been able to intake large amounts of data and analyze it much faster than a pre-automated era.
AI enables the privacy team to determine “the next analytical steps I need to take in order to render a decision or an opinion; whether this component and this component can share data with each other,” said Linares. It is his “personal goal” to use AI to build “knowledge links,” which he says is particularly useful when approaching a cybersecurity breach situation, where data is scattered and potentially in many different locations. An AI tool could help gather and sort that data faster in such a situation, when time is of the essence.
Wallace concluded the discussion by adding that “in the next few weeks” there will be a request for information issued by DISA regarding “how we can better leverage AI with respect to cyber defense.”
“And I know that’s asking you to boil the ocean,” Wallace admitted, “We’re not necessarily looking for boiling the ocean, but it’s what is the most impactful use of AI across the board to better help our defenders and better help the department defend itself.”
Continue the cybersecurity conversation with some of the federal government’s top cyber officials at the Potomac Officers Club’s 2024 Cyber Summit on June 6. All of the prominent GovCon companies will be in attendance, so send your best and brightest to conversate on this crucial topic. Browse the full lineup here and register here!
