Pentagon harnesses AI to revolutionise threat prediction
Artificial intelligence aids the Pentagon not only in identifying targets but also in predicting threats. The modern Maven system continuously analyses vast amounts of satellite data, thereby supporting the US military.
The Pentagon is investing in the latest technologies, including artificial intelligence. To this end, a special programme named Maven was created, which aims for in-depth analysis and prediction of dangerous situations. Vice Admiral Frank Whitworth, head of the National Geospatial-Intelligence Agency (NGA), emphasises that the biggest challenge is detecting unknown threats.
USA develops the Maven project
According to Defence One, the Maven programme is an initiative of the US Air Force and plays a key role in analysing satellite data. Work on the programme began back in 2017 and from the very beginning, the algorithms were trained to gather and analyse vast amounts of data. AI allows for the quick and efficient detection of troubling objects —enemy vehicles or changes in military object activities. The project has been dynamically developed over eight years, making it currently used by services and combat commands.
Artificial intelligence supports military systems
The NGA makes every effort to improve AI models that analyse data and provide information on enemy movements. Simply identifying objects is no longer sufficient. Currently, it is most important for systems to provide information along with in-depth analysis. Introducing such models will allow for precise threat prediction. At the same time, the Pentagon is working on the ASPEN system, which aims to help manage the growing amount of GEOINT data.
Increasing trust in AI
Militaries around the world are gradually gaining trust in artificial intelligence, realising its potential in increasing operational efficiency, intelligence data analysis, and threat prediction. The number of Maven programme users among soldiers and officers has quadrupled over the past year. Whitworth emphasises that with better models and appropriate computational power, the time to detect potential targets has decreased by 80%. Ensuring accuracy and reliability in threat identification is paramount.
Despite the successes, the NGA must be prepared for the growing demand for intelligence data. Experts warn that if proper steps are not taken, the demand for computational power could exceed available resources.
