Cyber Intelligence Initiative

How Artificial Intelligence and Quantum Computing are Evolving Cyber Warfare

Abstract

In today’s technological age, there are many means by which one state can attack another state. The attack could be, not only by conventional military, economic, or political methods, but also through cyber operations. Due to the advancement of technology and the increase in innovation, cyberspace is another battlefield of boundless threats. The use of technology against another state is known as cyber warfare, and it poses critical risks to national security. Cyber warfare can involve an enemy state hacking or disabling another state’s critical infrastructure systems or tapping into an intelligence database to take and gain valuable information. Cyber warfare has become an existential threat to national security. As the advancements of technology grow, so will the rate of cyber warfare attacks. Artificial intelligence (AI) and quantum computing (QC) are two great enhancers of the cyber domain. Due to their capabilities, they have a substantial impact on cyber warfare, but could have an adverse effect and significantly increase the number and threat level of cyber-attacks in the future. It is in the best interest of the United States to take the lead in both functional areas of cyber warfare to establish global dominance within the digital domain. If the United States can establish dominance in artificial intelligence and quantum computing, it would greatly influence the technological playing field in which we are engaged.

 

How Artificial Intelligence and Quantum Computing are Evolving Cyber Warfare

Cyber warfare is a digital attack orchestrated by a state or government with the intention of damaging computer systems and networks, committing acts of espionage, or mangling the critical infrastructure of an adversary or ally. Society’s increased dependence on technology in recent years has fostered a dramatic increase in the number of cyber warfare attacks. The first documented cyberattack was the Morris Worm, which hit an estimated six thousand computers. The cost of damages from this cyberattack stretched from one hundred thousand dollars to the millions (FBI, 2018). The Morris Worm only affected computers that ran a specific version of the Unix Operating System. The worm was attributed with seriously delaying computer functions and emails and destroyed computers in the process. Another example is the Stuxnet Worm that affected the Iran Nuclear Program, in which the United States was involved. Stuxnet targeted Windows computer systems and networks and was able to destroy close to a thousand uranium enrichment centrifuges. The result was slowing down the Iranian nuclear weapons program. Stuxnet was able to breach and replicate critical information by spreading very rapidly. The Stuxnet Worm was encapsulated in a USB drive, and, when the USB drive was inserted into a computer, it began its destruction at the nuclear weapons facility (M. Holloway, 2015). Current Director of National Intelligence Daniel Coats proclaimed how cyber warfare poses a critical threat to national security by saying, “Persistent and disruptive cyber operations will continue against the United States and our European allies, using elections as opportunities to undermine democracy, sow discord and undermine our values. Some of these actors, including Russia, are likely to pursue even more aggressive cyberattacks with the intent of degrading our democratic values and weakening our alliances. Iran will try to penetrate U.S. and allied networks for espionage and lay the groundwork for future cyberattacks, and North Korea will continue to use cyber operations to raise funds, launch attacks and gather intelligence against the United States.”

Artificial intelligence and quantum computing have both demonstrated their substantial prowess on the impact of technologies and the evolution of warfare as we know it. Artificial intelligence can collect excessively large amounts of data through algorithms that operate artificial intelligence systems. These algorithms can allow artificial intelligence to learn other abilities from rules or patterns in the data that construct the algorithms. Although very advanced in its capabilities, the danger of artificial intelligence being used against us, and therefore creating a vulnerability, is more than conceivable. Quantum computing has the ability to use quantum mechanical engineering to be able to perform simultaneous computations. A quantum computer is faster and more efficient than any computer known. Quantum computing poses a threat because, in theory, a single quantum computer would be more powerful than all of the supercomputers in the world today (J. Garamone, 2018). Continuing in theory, if quantum computing was fully mastered by a state, they would have the ability to hack networks, databases, and critical infrastructure systems with little to no resistance at all. Artificial intelligence and quantum computing both have the potential to become critically dangerous to national security due to the exceedingly difficult nature of defending against them.

 

Artificial Intelligence

Artificial intelligence (AI), otherwise known as “machine intelligence,” is a computer science division of intelligence that demonstrates the ability for humans to create technology that can enable computers or other forms of technology to become autonomous in their work, therefore mimicking human intelligence. AI has the ability to develop extensive problem solving, planning, research, and speech recognition skills (Allen, Chan, 2018). The algorithms that create AI are very complex and have completely surpassed the ceiling of its expectations, so much so that many experts fear how large and powerful AI will grow. AI systems have shown their uniqueness in several areas by proving more capable and efficient than humans today. There are many global companies that implement AI, including IBM and Microsoft. IBM’s Watson and Microsoft’s Azure Machine Learning Program are both able to perform computing tasks much faster and are able to conduct deeper and more thorough research than humans (Allen, Chan, 2018). AI has proven how beneficial and innovative it can be. On the international stage, however, AI’s greatest capabilities can, in fact, be its own greatest vulnerability. This poses potential risk when cyberwarfare is introduced.

There are many dangers that are associated with AI, including cyberattacks on existing AI systems, implementation of AI in conventional military warfare, and a greater overall threat to our national security. Current AI systems have begun to see data breaches from unknown sources due to unsecure centralized servers that hold valuable information. This creates an easy target for even the simplest of hackers to obtain information within these databases in bulk. Cyberattacks on AI databases could cause severe destruction for individuals, businesses, and our government (R. Browne, 2015). AI can also bring large technological advances to conventional military warfare, including the implementation of weapons systems that can have fully autonomous features and technology that is able to use complex problem solving and reasoning skills almost like humans. This can lead to an apocalyptic future dealing with AI weapons and possibly AI soldiers (R. Brown, 2015). Bryan McVeigh, a project manager with the Army, has publicly stated: “within five years, I have no doubt there will be robots in every Army formation.” The Pentagon has speculated they will spend nearly one billion dollars in the development of AI robotics (R. Browne, 2018). The idea of having robotics in warfare has influenced many founders of robotics and AI to postulate the danger that AI poses in warfare. They have urged the United Nations to ban lethal AI autonomous weapons, by saying, “Once developed, they will permit armed conflict to be fought at a scale greater than ever, and at timescales faster than humans can comprehend.” The ability for AI databases to be breached poses an immense national security threat as well. A cyberattack composed on a top secret database or on a power grid database could lead to severe critical infrastructure damages and mass human casualties. Comparatively, a massive cyberattack towards the United States could potentially cause more damage than the use of hard force via conventional weapon methods.

The integration of AI in the cyber domain has introduced challenges on the international front that has already affected international relations and will continue to do so. The tense relationships between the United States, China, and Russia has increased due to developments in AI technology. In July of 2017, China released details of its “New Generation AI Development Plan,” which explained China’s goals to advance AI technology (Hass, Balin, 2019). These goals include China catching up on AI applications and technology by 2020, achieving “major” breakthroughs by 2025, and becoming a global leader in AI by 2030. The United States’ response has been diplomatic, with the intention of slowing down China’s technological advances. To achieve this, the United States has applied targeted tariffs, increased the number of economic espionage prosecutions of Chinese agents, and placed a greater focus on counterintelligence operations and research (Hass, Balin, 2019). AI has also been used as a disruptive tool in political affairs, furthering tension between U.S.-Russia relations. The 2016 Presidential election showcased vulnerabilities within the United States’ election system of which Russia took advantage and interfered with the Presidential election. Using AI, Russia was able to spread misinformation with the intention of suppressing voter turnout in well-targeted geographic districts, as well as targeting groups by their demographics. Despite the dangers and challenges in AI’s vulnerabilities, the capabilities and advantages AI brings to the table fuel further research.

The evolution of AI technology has brought desirable advantages not only for the United States, but other nations as well. If the United States could position itself by being a global superpower in the field of AI technology, it could enjoy the advantages of AI technology which could aid in protecting from and mitigating future cyberattacks towards the homeland, as well as supporting intelligence agencies’ efforts. AI can be programed to protect, and, in turn, mitigate cataclysmal attacks. AI’s defense system, if created and managed correctly, could help prevent an attack on the homeland by another actor. A computer system or network using artificial intelligence would be able to defend against an attack more effectively than human effort alone, considering the more advanced technology than what is currently being used. Artificial Intelligence could also greatly impact United States intelligence agencies by enhancing capabilities for the analysis and collection of data, as well as helping in satellite imagery analysis and defending against other acts of cyber warfare (Hass, Balin, 2019). The underlying vulnerabilities and the desirable capabilities of AI reveal just why AI technology continues to evolve and be such a prominent tool for cyber warfare discussion.

 

Artificial Intelligence and Cyberwarfare

The relationship of AI and cyber warfare has changed drastically as a direct result of the digital age. Cyber warfare, with the use of AI, has become easier due to technological advancements, which allow a country to reach beyond its national and international borders. The primary cause of cyberattacks can be traced back to secured software system errors rather than other causes such as hardware failure. According to research conducted by Soorena Merat and Wahab Almuhtadi, the authors of “Cyber-Awareness Improvement Using Intelligence Techniques,” repeated occurrences of task management errors were discovered which cause the soft process to choose the wrong thread. This type of malfunction has detrimental effects on numerous computer programs and functions such as firewall and security programs. The devastation these type of attacks can cause represent an ongoing threat to cyber security in the United States.

When an error occurs and many threads have been mistakenly interrupted, oftentimes someone will try to manage this issue through a queue strategy. The importance of this process is the capacity to effectively manage multiple concurrent threads, while simultaneously performing high traffic tasks, like data transfers. One of the biggest concerns are delays while maintaining security. In order to ensure minimal delays, the process has a requirement called an “efficient feedback control.” The purpose of the efficient feedback control is to adapt the idle time to learn of changes in traffic dynamics by the constraint of queue length at a predetermined value.

In this article, which described the experiment performed by authors Soorena Merat and Wahab Almuhtadi, a sample data transfer process was utilized under the Destination-Sequenced Distance Vector (DSDV) routing protocol test without a firewall. The reason for the exclusion of a firewall is due to either a passive or active attack that can be initiated inside the security perimeter. At the center of their research is a program which implements the DSDV routing protocol used for the transfer, over a path, between nodes, in addition to handling changes in the routing table, which includes link break. This program has three threads; the first is for sending, the second is for receiving messages, and the third is for the main program, which then creates two other threads. The functionality of these threads can help the program identify and protect against cyber security attacks, whether they are passive or active. The sending thread checks batch files and updates the routing table of the node, transitions it to a message, and then broadcasts it. Additionally, the sending thread increases the sequence number of this node every few seconds. Meanwhile, the receiving thread continuously waits for messages from other nodes and updates the routing table once a message is received. Another component of the program is a lock mechanism to make sure only one thread would update the routing table, one time. The Merat/Almuhtadi experiment elevates and furthers the discussion of new technologies in cyber warfare.

 

Quantum Computing

Quantum mechanics is relatively old in the information age and has been gaining traction in the cyber warfare discussion due to its potential capabilities when a quantum computer is built. Quantum mechanics has evolved since the early twentieth century and has continued to build on previous discoveries. QC and its applications have proven to be incredibly complex as technology continues to advance.  In 1982, Richard Feynman proposed the idea of a quantum computer due to the impossible nature of executing quantum mechanical functions on a classic computer. David Deutsch, put Feynman’s theory into practice in 1985, and, by the 90’s, it was agreed that quantum computers would be exponentially faster than classic computers. In 1994, Shor’s algorithm revealed that quantum computers could factor large integers efficiently by using entanglement of qubits and superposition. Shor’s algorithm exercised the ability to break public key cryptosystems. The public key is the underpinning of internet security and the end game in the quantum arms race. In 1996, Grover’s algorithm discovered a square root speedup to find many applications to search and optimize problems. Also, in 1996, Seth Lloyd built on Feynman’s theory that a quantum computer was not only needed to perform quantum mechanics, but in fact could perform steps simultaneously, which classic computers could not do (Montanaro, 2017).

A quantum computer is defined by the National Academy of Sciences as a physical system that comprises a collection of qubits that are controlled and manipulated to implement algorithms to answer problems with high probability (National Academy of Sciences, 2018). Quantum computers are different from a classic computer in many ways, but the biggest difference is the bits on which it operates. A classic computer operates on binary digits or bits, which are combinations of zeros and ones that are manipulated and transmitted one step at a time to perform the functions we depend on daily. Contrarily, quantum computers operate on quantum bits (qubits). Qubits are like bits in that they can be zeros and ones, but, unlike classical bits, they can be in both at the same time. Qubits work simultaneously, giving an advantage over classical computers because they can reach solutions much faster. A quantum computer uses QC principles that classic computers are not capable of; superposition, measurement, collapse, entanglement, and coherence have been revealed over years of scientific discovery in the twentieth century. Superposition is the ability of qubits to exist in multiple states at once and is often referred to as the “superposition state.” Measurement is the observation of a quantum object interacting with a larger system, but, when measured, will yield a defined value, zero or one. This process is called collapse. The superposition collapses when you try to measure it. Entanglement is the physical interaction of particles in the quantum state, the building block of QC’s potential capabilities. Lastly, coherence is needed to shield any interaction from the outside, because even the slightest interference from the outside environment will cause quantum properties to decohere. These unique properties set quantum computers apart from classical computers and will pilot more discoveries in the areas of encoding, manipulating, and the transferring of information (National Academy of Sciences, 2018).

Although quantum computers have immense capabilities, they also possess a great deal of self-imposed constraints within their own systems. The National Academy of Science indicates four constraints that limit the power and capabilities of quantum computing. The first is the lack of interaction of entangled qubits, since entanglement is one of the core properties to have quantum computing power. The second constraint is the no-cloning principle. The inability to make and store a copy, as classical computers can, requires new algorithm approaches. The third constraint is the incapacity to block out outside environmental noise that could detrimentally affect the advantages quantum computing possesses. The last constraint the Academy of Science noted is the inability adequately to measure particles in a quantum state. An algorithm that can manipulate the system to produce the desired solution is needed (National Academy of Sciences, 2018). Despite its constraints, the potential power alone is motivating nations around the world to develop the first large scale, working quantum computer.

Many of the dangers of QC are unknown because the progression of its technology is so slow. One thing we do know is that QC’s will have the capability to break public key cryptosystems, which appears to be most valuable. The competition between nations to develop the first functional QC has been observed as the arms race of the technological era. Currently, the U.S. is competing with China, Russia, Canada, Japan, Israel, and Europe to exert its dominance in the digital domain (Friedson, 2017). Scientists from National Institute for Standard Technology (NIST) report, “If large scale quantum computers are ever built, they will be able to break many of the public key cryptosystems currently in use.” This completely revolutionizes warfare, as it creates new resources for those who are able to develop it in time. Depending on which nation can create a working quantum computer with access to other nation’s public keys, they will hold the power to disrupt militaries, government infrastructure, and economies. The nation that comes out on top will exert dominance and have the power to change the political landscape in the future (Friedson, 2017). If it falls into the hands of other nations, the ability to hack into public keys is mutual assured destruction. This furthers the comparison of the QC arms race with the nuclear arms race. That is why the dominance in the race to create a quantum computer is paramount.

 

Quantum Computing and Cyberwarfare

Quantum computing is a potential threat to cyber security. The vulnerabilities of code-breaking and public-key cryptography leave the United States susceptible to cyber terrorism threats. An article in the February 2019 American Scientist, written by Dorothy E. Denning, is titled “Is Quantum Computing a Cybersecurity Threat?” Denning expresses concerns that, with continuous technological advancements, the processing abilities of computers may be able to break encryption keys in the future. During her research, she found that a new type of computer, based on quantum physics, could break modern day cytopathy, leaving communications insecure as if they were never encoded, causing international devastation. However, this is completely hypothetical of the worst-case scenario. Today, all current quantum computers do not have the processing capabilities to carry out such a massive threat. Fortunately, for any country or cyber terrorist organization to use quantum computing in this manner, drastic and substantial technological advancements would be required.

 

AI and QC Cybersecurity

The evolution of AI and QC in modern warfare leads to security implications as well. Advancements in the digital age which have made our greatest strengths possible may also be our greatest vulnerabilities. As a nation that wants to lead the world in the digital domain, we must work on securing our cyberspace and make it resilient to future attacks. While AI is being used as weaponry, to teach machines, and to program drones and autonomous vehicles to do harm instead of good, it can be disguised and not thought of as a tool of destruction (Cetron, Davies, 2009). Nicole Eagan, CEO of cybersecurity at Darktrace said, “We’re still in the early days of the attackers using artificial intelligence themselves, but that day is going to come…And I think once that switch is flipped on, there’s going to be no turning back, so we are very concerned about the use of AI by the attackers in many ways because they could try to use AI to blend into the background of these networks.” Added layers of security is of the utmost importance when considering QC and the race to develop a quantum computer. For the United States to establish dominance, and protect itself in the process, it must begin to establish precautions to create a security system and increase technical training and cooperation with the private sector. According to NIST, the United States, “should begin now to prepare our information security systems to be able to resist quantum computing.” This entails protecting the computers citizens use every day to protecting the critical infrastructure we rely on to go about our daily lives. We must promote and incentivize new technical training in cyber security for individuals and companies to protect their computer systems. The U.S. government needs to increase research budgets to continue evolving technologies. According to the World Future Society, the United States produces fewer scientists, engineers, doctors, and technicians than China and India. Going into the technological battlefield, that puts the United states at a disadvantage (Cetron, Davies, 2009). Being preemptive as well as continuing the development of AI and QC will convey to the rest of the world that the United States will be a leading player until it achieves dominance.

Conclusion and Future Study

The influence of Artificial Intelligence and Quantum Computing will continue to be a leading topic in cyber warfare discussions due to the immense capabilities and uncertain vulnerabilities in these two fields. AI and QC can be utilized as a great advantage within the new battlefield of cyberspace, and, without proper protection, they both can aid in our own destruction as well. In order to exert global dominance in the digital domain, we must be a leading nation in the areas of AI and QC. With that offense, we also must ensure a proper defense and continue to ensure our cyber space is secure and resilient against cyberattacks. It is in the national interest of the United States to become a leader in the realm of AI and QC, because, by doing so, it can better protect the homeland from cyber warfare attacks. To further the research of this paper, a deeper look at the leading nations in competition to establish dominance in the digital domain will be studied. An in-depth study of current capabilities of the United States, China, Russia, and Iran will be analyzed, compared, and contrasted to predict what nation is taking the lead.


References

Allen, G., & Chan, T. (2018, June 28). “Artificial Intelligence and National Security.” Retrieved from https://thebulletin.org/2018/02/artificial-intelligence-and-national-security/

Bachman, J. (2018, May 18). “US Army Turns to Robot Soldiers.” Retrieved from https://www.bloomberg.com/news/articles/2018-05-18/the-u-s-army-is-turning-to-robot-soldiers

Browne, R. (2018, July 25). “Weaponized drones. Machines that attack on their own. ‘That day is going to come’.” Retrieved from https://www.cnbc.com/2018/07/20/ai-cyberattacks-artificial-intelligence-threatens-cybersecurity.html

Cetron, M. J., & Davies, O. (n.d.). “World War 3.0: Ten Critical trends for Cybersecurity” (Vol. 43, Ser. 5). Washington, D.C.: The Futurist. Retrieved February 23, 2019, from www.wfs.org.

Denning, D. E. (2019). “Is Quantum Computing a Cybersecurity Threat?” American Scientist, 107(2), 83-85. https://doi.org/10.1511/2019.107.2.83

FBI. (2018, November 02). “The Morris Worm.” Retrieved from https://www.fbi.gov/news/stories/morris-worm-30-years-since-first-major-attack-on-internet-110218

Friedson, Idalia. (2017, May 02). “The Quantum Computer Revolution Is Closer Than You May Think.” National Review. Retrieved March 11, 2019, from https://www.nationalreview.com/2017/05/quantum-computing-race-america/

Garamone, J. (2018, February 13). “Cyber Tops List of Threats to U.S., Director of National Intelligence.” Retrieved from https://dod.defense.gov/News/Article/Article/1440838/cyber-tops-list-of-threats-to-us-director-of-national-intelligence-says/

Hass, R., & Balin, Z. (2019, January 10). “US-China relations in the age of artificial intelligence.” Retrieved from https://www.brookings.edu/research/us-china-relations-in-the-age-of-artificial-intelligence/

Holloway, M. (2015, July 16). “Stuxnet Worm Attack on Iranian Nuclear Facilities.” Retrieved from http://large.stanford.edu/courses/2015/ph241/holloway1/

Merat, S., & Almuhtadi, W. (2015). “Cyber-Awareness Improvement Using Artificial Intelligence Techniques.” International Journal on Smart Sensing and Intelligent Systems,8(1), 620-636. doi:10.21307/ijssis-2017-775

Montanaro, A. (2015, November 25). “The Past, Present, and Future History of Quantum Computing.” Retrieved from https://people.maths.bris.ac.uk/~csxam/teaching/history.pdf

National Academies of Sciences, Engineering, and Medicine 2018. Quantum Computing: Progress and Prospects. Washington, DC: The National Academies Press. https://doi.org/10.17226/25196.