Kissimmee, Fla. – As U.S. intelligence companies improve their reliance on industrial geospatial services, cybersecurity turns into a rising concern.
“We create so many progressive options right here on this nation,” Stacey Dixon, U.S. deputy director of nationwide intelligence, mentioned Could 6 on the 2024 GEOINT Symposium right here. “But we lose a number of that innovation to adversaries as a result of we aren’t correctly defending it. Spend money on your cybersecurity. Perceive that persons are going to attempt to steal your secrets and techniques.”
Potential adversaries will search entry to industrial merchandise immediately in addition to by means of intermediaries.
Work out methods to stop that, “as a result of we don’t need our American-made merchandise or our western-made merchandise falling within the fingers of adversaries who will use them in opposition to us, use them in opposition to our troops, use them in opposition to our pursuits,” Dixon mentioned. “Please take the time to take a position and know who’s shopping for these capabilities.”
Synthetic Intelligence
Data safety can also be important as authorities companies flip to synthetic intelligence for assist combining myriad datasets.
“If we’ve got algorithms that may get into the datasets which are presently held in numerous organizations and companies and run all of that information on the identical time, it’s going to be more practical and supply totally different sorts of insights than if you happen to do this stuff in parallel after which mix them on the finish,” Dixon mentioned.
Creating these algorithms shall be a problem. As well as, U.S. intelligence companies searching for to mix datasets with these of different companies could have to undertake new insurance policies. Laws may be required.
“We had been created as a group with specific companies with specific disciplines,” Dixon mentioned.
Legal guidelines describe what sort of information every company acquires.
“We’re going to have to determine methods to work by means of that to nonetheless defend the American public whereas enabling extra individuals to have entry to the info,” Dixon mentioned.
Confirm and Validate
Earlier than intelligence companies can belief algorithms and automation, they might want to perceive how machines are developing with the knowledge, Dixon mentioned. For giant-language fashions, for instance, “let’s be capable of hint it again. That’s doable.”
Given growing misinformation and disinformation campaigns, coaching analysts to confirm and validate data sources can even be important, Dixon mentioned.
Analysts who develop forecasting instruments, for instance, might want to show the worth of the instruments.
One of the best ways to check a forecasting software “is so that you can inform me one thing that’s going to occur sooner or later,” Dixon mentioned. “Then let’s to see whether or not your prediction was right.”
When fashions show themselves, analysts will undertake them as a result of the fashions will cut back the period of time analysts should dedicate to mundane, repetitive duties, Dixon mentioned.
“It permits them to make use of their artistic brains to enhance what the machines are telling them,” she added.
