Integration of Artificial Intelligence into the US Intelligence System: An Analysis of the Current Situation of Artificial Intelligence Enhancing Open Source Intelligence

The U.S. intelligence community collects and analyzes a large amount of intelligence data every day, which not only serves the combat missions of U.S. troops stationed abroad, but is also used to combat crime, drug trade and terrorism at home. The intelligence community has been actively developing technical tools for a long time, trying to use artificial intelligence technology to improve the analysis process and help analysts conduct intelligence analysis more effectively, which also reflects good application practices in intelligence work.

Analysis of the Current Situation of Artificial Intelligence Enhanced Open Source Intelligence

1. Integrate massive data information

The disorder, complexity, and huge quantity of open source information have become the main limiting factors for the development of open source intelligence means in the information age. The essence is that it is difficult to accurately capture and screen open source information. Currently, U.S. intelligence agencies and the military are increasingly using machine learning-based analytics platforms to sift through data sources such as social media. Its purpose is to realize automatic detection, improve the efficiency of grabbing and screening data, and finally collect a large number of documents that can be used for analysis and automatically generate standardized materials including tasks, locations and times. According to a survey in the United States, in 2019, an average of more than 4 million hours of video content was uploaded to YouTube every day, and about 4.3 billion messages were posted on Facebook. In the face of massive data, only relying on intelligence personnel to process and analyze it can no longer meet the demand. In contrast, artificial intelligence can break through the calculation speed and endurance limitations of the human brain, greatly improve the efficiency and stability of intelligence data processing, and meet the processing capabilities required by open source intelligence.

Electrif Ai, a defense contractor that provides services to the intelligence community, has introduced an algorithm called Signal Sensor. The algorithm can use machine learning to detect and identify threats on websites such as social networking platforms and online forums with huge amounts of data, and can rank targets according to the severity of threats. AI is already intertwined with U.S. open-source intelligence and traditional military domains. In the entire intelligence cycle, there are a large number of artificial intelligence-driven physical drones and sensors, as well as a large number of web crawlers and spider programs in cyberspace. They continue to collect a large amount of data and form an automatic and semi-intelligent communication mechanism for the target, realizing the role of artificial intelligence in accelerating the intelligence cycle.

2. Improve the availability of open source intelligence

The Congressional "Artificial Intelligence and National Security" report argues that artificial intelligence systems may have the ability to integrate large amounts of data from multiple sources. In the past, the task of open source intelligence work was to find hidden or hard-to-find information; now, facing the increasing amount of data, another challenge is to mine the knowledge behind a large amount of information. In October 2019, the U.S. Army Intelligence and Security Command (INSCOM) signed a $437 million procurement contract with BAE Systems. It is designed to provide open source intelligence support to U.S. Army troops stationed abroad, providing regional commanders with early warnings of riots, political violence, terrorist attacks, or other security-related incidents. Similarly, the Stabilitas company funded by the CIA also uses automatic capture of network information and monitors public opinion and regional situation in a certain area through sentiment analysis. Stabilitas' artificial intelligence program continuously draws from more than 17,000 global data sources and identifies nearly 300,000 critical events every day, such as natural disasters and geopolitical events. Afterwards, the company used software identification and analysis to find out the interconnection between the people and assets affected by these events, so that the CIA can effectively monitor sudden natural disasters and mass incidents around the world.

Overall, the development of artificial intelligence in OSINT can be roughly divided into three stages.

The first stage is used as a tool for data collation and screening.

The second stage is human-machine collaboration, that is, the automatic capture of open source data and the semi-intelligent analysis mechanism under human intervention.

The third stage is autonomous operation, that is, artificial intelligence repeatedly trains algorithms based on massive historical data, and finally realizes relatively independent independent and accurate collection and analysis.

At present, the main development direction of artificial intelligence in American open source intelligence work is still human-computer cooperation, and it is still very difficult to realize artificial intelligence technology that is completely free from human intervention. In view of the double complexity of open source information data format and content, the use of technology to enhance human cognitive ability is still the goal of artificial intelligence development in open source intelligence work.