Early AI use cases for the Department of Defense
The Department of Defense is maturing AI models for predictive maintenance with reinforcement learning to catch equipment failures before they occur and perform proactive maintenance. Ultimately, there will be no need for humans to review sensor data from aircraft, for example. AI will be able to recommend when planes should be taken out of service and how many more hours of flight time can be tolerated before repairs.
The AI has become even more granular and can inform mechanics of the likelihood of a component failing within a set time frame after a series of events. Such analysis paves the way for self-healing, where AI automates remediation before problems become manifest.
In the cybersecurity space, the military is interested in leveraging AI to identify insider threats and zero-day vulnerabilities faster than inexperienced, underpaid contractors who tend to rely on scrutinizing log data. . The goal is for the system to be able to use news feeds to identify vulnerabilities exploited by foreign adversaries such as Iran, search for them across networks, and suggest security controls.
There are also AI applications for autonomous unmanned aerial vehicles. A UAV can also be placed on the target to allow the model to process ground conditions and recommend course of action. In another model, the drone could start monitoring mobile traffic when the subject picks up the phone. With current methods, human-piloted drones collect all the data all the time, but it’s too much information for analysts to digest.
AI could be used for geospatial analysis of imagery to assess building damage and landscapes in unknown areas. The Pentagon also wants AI to be able to compare images and determine if the aircraft made any moves that would indicate takeoff or landing. The human eye might miss it, but a machine that analyzes the image pixel by pixel would miss it.
read more: Watch expert discussions about what AI is and the benefits of the technology.
Private institutions also recognize the need for AI
Healthcare organizations find it easier for AI to catch Medicaid fraud than humans because breadcrumbs are often buried within disconnected data sets.
Similarly, AI can analyze disconnected data sources to help government agencies make recommendations for rerouting resources in advance of a natural disaster.
The Department of Justice is using AI to predict situations that may require law enforcement presence. The challenge is the sheer volume of data that government agencies must parse. The National Security Agency’s Utah data center alone is growing by a petabyte each month.
While some agencies may want their websites to be detectable by ChatGPT or other generative AI models, others will no doubt be using third party software for profit. I have a problem with handing over data to These agencies may develop their own language models for tracking metrics such as public engagement with programs and services.
This article is part of fedtechof CapITal blog series.
