The huge volume of data brought about by the proliferation of sensing devices and information technology presents both opportunities and challenges for the US and Chinese intelligence communities, as both countries seek to incorporate the latest artificial intelligence (AI) technologies into their intelligence systems and enhance their spying capabilities. Analysts say that China’s military-civilian integration of technology, theft of large amounts of data through cyber intrusions puts pressure on U.S. intelligence efforts, the U.S. government must work more closely with private technology companies in the field of artificial intelligence.
Must navigate massive data waves, or drown.
The U.S. intelligence community believes that the rate at which openly available information is being generated is growing exponentially and has long ago exceeded the ability of intelligence systems to understand and respond. There are many reasons for this surge in data – they come from commercial sensors, covert sensors in the military and government, cyberbots (bots), open-source intelligence, drones, small satellites in space, and more – and the manpower of intelligence agencies has long been unable to cope with such massive amounts of data.
Susan Gordon, then deputy director of national intelligence, said in 2019: “To achieve the desire to ensure intelligence superiority, the U.S. intelligence system must adapt to the rapid global popularity of sensing, communications, computing and data machine analysis. These trends threaten to erode the previously unique capabilities and strengths of the U.S. intelligence system; going forward, we must improve our analytical capabilities and learn from the intelligence system’s large-scale data collection.”
Anthony Vinci, former chief technology officer and deputy director of the U.S. National Geospatial-Intelligence Agency (NGA), said in an article published on the Foreign Affairs website in August this year that artificial intelligence will bring a “new revolution” to global intelligence services. He said that in this intelligence revolution, the machine will not only be a tool for information collection and analysis, the machine will also become a direct user of intelligence, decision makers, and even rival machine intelligence work target.
The U.S. intelligence system includes 17 departments, including the National Geospatial-Intelligence Agency, which in 2017 acquired 12 million imagery feeds and tagged 50 million index points.
Then-Director Robert Cardillo predicted that the NGA would need to process a million times more imagery over the next 20 years, and that the government would have to hire 8 million imagery analysts to handle such astronomical numbers under traditional manual means.
Caldiero said in a speech that a single U.S. military sensor in a battlefield gets the equivalent amount of data per day as the amount of high-definition footage produced by all the games in the National Football League (NFL) over three seasons.
He said that in such massive amounts of data, “intelligence agencies either harness it …… or are overwhelmed by it.”
The “Three A’s” of the U.S. Intelligence System Strategy
The Office of Transformation and Innovation (OTI), part of the Office of the Director of National Intelligence (ODNI), on October 22, 2020, issued a Request for Information (RFI) to the private sector and non-governmental organizations (NGOs), seeking proposals for technology, research, and more to support strategic initiatives of the U.S. intelligence system. This is part of the Office of the Director of National Intelligence’s (ODNI) Intelligence, Science, and Technology Partnership Program (In-STeP), which seeks to provide the intelligence system with greater insight into technologies and business practices being developed by the private sector.
In particular, the invitation welcomes input from the U.S. private sector on how to use artificial intelligence to enhance government intelligence capabilities.
The Augmenting Intelligence using Machines (AIM) initiative, announced in January 2019 by the Office of the Director of National Intelligence (ODNI), identifies the intelligence system’s “three A’s” strategy – sustaining U.S. intelligence superiority through Artificial Intelligence, process automation, and augmentation technologies for intelligence community officials.
Matt Connor, former information security officer at the National Geospatial-Intelligence Agency, said the NGA has been using Artificial Intelligence Markup Language (AIML) for the detection of anomalous behavior in networks, and the advantage of AI is that it can clearly troubleshoot suspicious problems in networks amidst a lot of complex information.
“We’ve found that the only way to detect anomalous behavior in a system is if you do trend analysis on large data sets over time, which reveals something to explore,” Connor told the Federal News Network in an interview. “So we see the value of (AI in) network cleanup and monitoring for malicious content.”
The U.S. National Security Council on Artificial Intelligence (NSCAI) noted that as international competition for dominance in AI accelerates, the battlefield advantage will “shift to those countries with superior data, connectivity, computing power, and algorithms, a battlefield that will expand beyond the military to the intelligence field as AI and related technologies permeate intelligence operations.
Cybersecurity expert warns: Data stolen by Chinese hackers will be used for intelligence work
According to cybersecurity expert Nicholas Eftimiades, China’s multiple cyber intrusions against U.S. targets have obtained large amounts of personnel data, and artificial intelligence can help Chinese intelligence agencies effectively integrate the information and strengthen China’s ability to penetrate U.S. intelligence.
The US government has accused hackers working for the Chinese government of stealing 21.5 million highly sensitive military and government department personnel data from the US Office of Personnel Management (OPM) since 2013. The U.S. also alleges that Chinese hackers broke into several private sector user databases in the United States.
Efidimyades currently teaches at Pennsylvania State University and has worked for several federal agencies, including the US Department of Defense and the State Department, as well as the Central Intelligence Agency (CIA) as a Technical Operations Officer. He said: “They (the Chinese hackers) have access to the biographical data of all the employees who have reached a certain level of security. That allows them to go through that data and find out who has security clearances from which agencies, and by comparing the list of foreign diplomats with the list of other employees in different positions, they can see what kind of access they have, what kind of classified information they have, and what kind of organizations they’re working for. So this brings huge benefits to China’s counterintelligence efforts and intelligence targeting efforts,”
In February, the U.S. Department of Justice indicted four Chinese military hackers for hacking the U.S. credit reporting agency Equifax in 2017, stealing data on nearly 150 million people.
In 2018, U.S. hotel chain Marriott International announced that hackers had stolen the personal data of nearly 500 million customers from its Starwood reservation system starting four years ago. U.S. Attorney General Barr confirmed in February this year that this cyber intrusion was done by Chinese hackers.
In 2015, two Chinese hackers broke into the computer systems of Anthem, a U.S. health insurance company, and stole the personal data of at least 78 million Americans. The two hackers were indicted in 2019.
Yahoo News reported in 2018 that Chinese intelligence services successfully hacked into passenger biometric data at Bangkok’s Taikoo Airport. The report quoted a former U.S. intelligence official: “The Chinese have been extracting data from all the major transportation hubs in the world,”
Efidimyades said: “We found about six Chinese intrusions to obtain medical data. Some statistics show that China has personally identifiable information on 80 percent of the U.S. population.”
“This data is useless without artificial intelligence and large datasets to bring all of this together, so this is the way forward for China.” He told Voice of America.
Is China’s military-civilian integration an institutional advantage?
Brian Katz, a researcher on international security and transnational threats at the Center for Strategic and International Studies (CSIS), recently published a research brief stating that in the evolution to “smart” warfare, competitors such as China and Russia will enjoy a structural advantage: a structural advantage in the development and use of “smart” weapons. In terms of artificial intelligence technology, the route of military-civilian integration. This resource advantage will be used to strengthen defenses against U.S. intelligence operations and enable more targeted and aggressive offensive operations, he said.
The ability of China, Russia and other authoritarian states to integrate civilian and military AI research and development and direct commercial sector innovation toward military and intelligence applications allows them to pool national resources and know-how and potentially adapt more quickly to the changing operational environment, the brief said. China’s ongoing advances in 5G and IoT will enable faster deployment and use of AI-enabled intelligence tools, both defensive and offensive.
He also warned that cyber-attacks with AI technology could target collection and communication platforms and use intelligent malware to access, exploit or destroy critical data and intelligence. Once inside, foreign intelligence agencies could use “anti-artificial intelligence” techniques to insert “viruses” or fake data into “training sets” to confuse the artificial intelligence of United States intelligence agencies. algorithms, and lead to poor performance of AI systems, such as recognizing friends as enemies.
In addition, AI-backed disinformation campaigns will allow adversaries to spread disinformation on an unprecedented scale, creating confusion for analysts and policymakers trying to understand and act on the information.
Efidimyades said, “Although we have the AI Council and the Department of Defense also leading AI development efforts, most AI development is done outside of government, in companies and academic institutions. Because of our separation of government and business, we don’t have a national effort like China, like the Communist Party of China.”
China continues to strengthen the strategic position of AI and focus on the application of emerging technologies in the military and intelligence fields. For example, Nanjing Regent Intelligence Technology, established in 2018, works closely with the Chinese military and intelligence services. Public information shows that the company will integrate advanced algorithmic models with national defense scenarios to develop product solutions for military customers in the areas of “knowledge, intelligence, strategy and anti-depth counterfeiting,” and will also launch intelligent products and technology solutions in business areas such as intelligence and decision support.
The mainstream analysis of the science and technology sector believes that the United States has a deep foundation in basic research in artificial intelligence and has mastered the core technology, while China’s advantage lies in the application and commercialization layout. But the CIA’s former technical action officer Effie Di Miades believes that, compared with China, the U.S. AI field of “military-civilian integration” is not enough.
“While we are guiding the development of AI for government use at some levels, from healthcare to military to intelligence, we don’t have a national effort. We are just now bringing together the various elements at the national level to make demands for the development of AI.” He said, “The change in AI is right around the corner, and that’s why the U.S. government is refocusing and going all out to bring government and industry together.”
Efidimyades said, “The U.S. has been figuring it out, and it’s only in the last year or so that we’ve started to put these elements together. But we’re still behind.”
Recent Comments