Artificial Intelligence’s Impact on National Security

Abstract

Artificial Intelligence (AI) can produce or create rapid technological advancement when applied to research; because of this, AI has been widely adopted across industries. This widespread adoption presents both an opportunity and a challenge to national security. The growing advancements in AI technology necessitate the need for research and focus on the ethical issues surrounding AI. AI represents an evident national security risk, demanding immediate policy attention from national leaders to prevent potential threats. This paper provides an overview of AI, explores its challenges as a national security concern, and outlines essential mitigation policies that national leaders should prioritize.

Introduction

Artificial Intelligence (AI) is characterized by its ability to mimic human-like intelligence and perform complex tasks. Recently AI has gained a significant spotlight and has rapidly advanced in the past year, revolutionizing technology and prompting ethical dilemmas surrounding the advanced technology. Artificial Intelligence and Machine Learning have become "Buzz" words surrounding major industries. While AI offers immense opportunities for progress, it also presents unprecedented ethical challenges, necessitating effective governance and protection against potential security risks stemming from AI. Research focused on AI and its intended impacts is of paramount importance. In today's rapidly evolving technological landscape, AI research significantly impacts U.S. National Security. Artificial Intelligence offers rapid advancements in many sectors and industries in the United States. However, due to the nature of its intelligence, AI poses an unprecedented ethical conundrum due to the potential security risks and challenges surrounding governing and protecting the United States from the massive capabilities and potentials stemming from AI.

Research in the field of AI provides valuable insights into the societal, economic, and security implications of AI, guiding policymakers, industry leaders, and researchers in making informed decisions and formulating effective regulations. By examining the intended impacts of AI, research contributes to the development of ethical frameworks, safeguards against misuse, and strategies to harness the transformative potential of AI while mitigating its potential risks. National security efforts must focus on mitigating AI threats, and technological development surrounding AI needs to be monitored and regulated.

AI Revolution in National Security

Cutting-edge technologies such as telecommunications or nuclear technologies have always prompted a discussion of ethics. Fully understanding the technologies leads to formulating the technology's ethical use and boundaries. In many regards, AI technology has defense applications used by the military. Research and the understanding of advanced technologies are imperative surrounding their use. Regrettably, AI possesses a significant level of risk, comparable to previous cases and potentially even more precarious due to the rapid pace of technological advancement and the complex interplay between government and industry (Allen & Chan, 2017). Many AI researchers have stated they do not fully understand how machine learning and the algorithms related to AI begin to perform deep learning, coining the phrase "black box." – this term means there is essentially a black hole in the knowledge. At some point, the models begin learning in a depth that the researchers do not understand the thought process of the deep learning systems.

"Black Box of AI"

AI research reaching this point of “black box,” questioning how AI directly works within a programed environment is a cutting-edge topic of discussion. AI has made notable changes and advancements to reach its current point of impact. AI will also continue to have broad impacts on society before relevant ethical problems arise stemming from its impacts (Amodei et al., 2016). A starting point worth noting is that in 2011 AI had one of its first victories as IBM’s Watson Won Jeopardy – which demonstrated the ability for AI to achieve natural language processing. AI’s abilities have continuously and rapidly changed over the past ten years to the most recent years. AI has revolutionized the national security landscape as it has impacted data analytics, surveillance, advanced fighting capabilities, and weaponry.

AI Dark Side – Ethics, Challenges, and Risk

            One of the notable changes in AI is the unprecedented progress in data analytics. AI algorithms have become more sophisticated and efficient in processing vast amounts of structured and unstructured data, enabling organizations to extract valuable insights and patterns for decision-making. This has revolutionized intelligence analysis, enabling analysts to detect trends, identify potential threats, and make informed assessments more rapidly and accurately. AI has also evolved computer "vision" leading to significant advancements in surveillance capabilities. Advanced image and video analysis algorithms now allow for real-time object recognition, facial identification, and anomaly detection, enhancing the ability to monitor and secure critical infrastructure, borders, and public spaces. AI-powered surveillance systems have become increasingly intelligent and accurate, supporting security operations with actionable information and early warning mechanisms. AI’s advancement proves its progression in terms of having a mental state. AI is being programmed to solve challenges faster than human beings. This brings up the ethical obstacles we face with AI, consider this quote by AI researchers, "The prospect of AIs with superhuman intelligence and superhuman abilities presents us with the extraordinary challenge of stating an algorithm that outputs superethical behavior" (Bostrom & Yudkowsky, 2014).

AI has also had a major role in Autonomous systems, being applied to military surveillance, precision targeting, and autonomous decision-making. Unmanned aerial vehicles, unmanned ground vehicles, and autonomous weapons systems are all examples of AI-enabled technologies that have transformed the landscape of modern warfare (Work, 2021). The cybersecurity domain has also experienced the impact of AI advancements—advanced threat detection and response systems, capable of identifying patterns of malicious activities, detecting anomalies, and responding in real-time to cyber threats is one of the positive impacts of AI. AI-powered cybersecurity tools play a critical role in safeguarding sensitive information, critical infrastructure, and networks from cyberattacks, providing a proactive defense against evolving threats. AI also proves to be an area of threat as malware developed with AI can actively rewrite its code to remain undetectable by malware detection software. The same AI technology being used to positively impact industry also has the capability to enable the malicious uses of AI as the technologies mature (Brundage et al., 2018). These advancements have revolutionized national security practices by providing more sophisticated and efficient tools for intelligence analysis, surveillance, military operations, and cyber defense, but have also challenged national defense to face the challenges of defending against the malicious use of AI.

Responsible AI: Recommendations on how to Approach AI

Integrating AI technology into various aspects of society brings forth multiple ethical challenges that must be addressed. These ethical challenges share similarities with the integration of cellular technology into everyday life. These challenges involve issues such as privacy, bias, transparency, and accountability. As AI systems become intertwined into industry and everyday life, ensuring ethical behavior and preventing unintended consequences, have become critical concerns surrounding AI. Recent studies have highlighted the potential dangers of autonomous weaponry, where AI-powered systems can make critical decisions in military contexts. However, trusting AI to do so without continuous human oversight is an ethical dilemma.

Researchers have been exploring autonomous AI technology and are calling for Human oversight, “Today, AI systems can launch attacks and defend against them, with little or no human intervention. A race to develop autonomous weapons could be even more dangerous than the nuclear arms race because barriers to entry are lower. Even small nations and non-state actors could exploit the technology to develop bespoke, disruptive capabilities" (Taddeo & Floridi, 2018). The development and deployment of such technologies raise questions about human oversight, the potential for unintended harm, and the need for robust safeguards to prevent misuse or abuse. In many industries, human oversight should be considered mandated to safeguard against the potential threats of AI itself and become an industry standard to protect against AI's unknowns. The "black box" nature of AI refers to the challenge of understanding how AI systems arrive at their decisions. Deep learning algorithms can process vast amounts of data and extract complex patterns, but the inner workings of these algorithms and their complexities are difficult to interpret or explain. This lack of understanding creates a lack of transparency surrounding AI and inevitably raises concerns about accountability, fairness, and potential biases embedded within AI systems. Research efforts are underway to develop explainable AI methods and techniques to shed light on the decision-making processes of AI systems.

These unknowns being researched, related to the "black box," are directly a national security concern. AI technologies could be exploited for malicious purposes, such as orchestrating cyberattacks, spreading disinformation, or disrupting critical infrastructure (Scharre & Lamberth, 2022). Additionally, the unintended consequences stemming from the "black box" nature of AI could lead to unforeseen outcomes and unintended harm. It is crucial to establish robust governance mechanisms, international collaborations, and regulatory frameworks to address these risks and prevent the misuse of AI technologies. Doing so will help mitigate the potential for unintended consequences stemming from AI advancements. Governance, policy, and litigation will have to be put into place to mitigate potential risks of AI that stem from the lack of human control and the potential for autonomous AI-driven systems to violate international humanitarian laws or for their AI systems to become manipulated for malicious purposes.

The different administrations of Obama, Trump, and Biden have approached AI differently, addressing components of AI that were most pressing or concerning at the time. During Obama's time in office, his administration addressed the initial fostering and promotion of research surrounding AI—pushing for funding and government-related collaborative opportunities. During the Obama administration, the potential for AI was just beginning to be seen; again, this was right when IBM's Watson won Jeopardy. During Obama's administration, there was less of a concern surrounding AI's governance as its capabilities were not as developed as they are today. Under the Trump administration, the push for development was further promoted. The administration aimed to create a favorable environment for AI innovation and deployment, with an emphasis on economic competitiveness. Research in 2020 begins to point to ethical concerns due to the applicability of AI across various industries; this led to the call for more governance surrounding AI. The Biden administration has had to focus on creating a balance between governance and the promotion of AI development. Determining what the correct sweet-spot is to balance the development of AI to further national security and to protect national security from threats faced by AI development. As AI has developed dramatically between 2008 and now, we can see how the focus of administrations has transitioned with the research; as more and more projects begin to point to the immense power of AI, a focus on researching the containment and the full understanding of AI has increased.

Governance, policy, and litigation surrounding AI have been steering in a positive direction under current and past administrations. To continue the development of frameworks and governance surrounding AI, there are several key recommendations to be considered. There must be a promotion of interdisciplinary collaboration surrounding AI and transparency. It is vital to promote openness and transparency in the design and functioning of AI systems. This includes making the decision-making processes of AI algorithms understandable and providing clear explanations for AI-generated outcomes. Open design principles encourage collaboration, accountability, and trust among stakeholders. Pushing for open design and collaboration on AI will be a challenge, such as it is with Cybersecurity technology and how it was challenging to discuss nuclear technology and treaties. Open design and collaboration will address concerns such as privacy, bias, and accountability. Controlling and implementing "safe" AI practices is another key aspect. This can be achieved by adhering to the principles of the CIA triad: confidentiality, integrity, and availability. Confidentiality ensures data privacy and protection, integrity focuses on maintaining the accuracy and reliability of AI systems, and availability ensures that AI resources are accessible and operational. However, this is centered around open design: fundamentally, there must be an open and continuous discussion surrounding the research and potentials of AI. A strategic approach can be taken to current and emerging national security concerns surrounding AI by having continuous and open conversations.

Conclusion

Artificial Intelligence and its rapid advancements pose an ongoing threat to national security. AI has greatly developed over the past decade and will continue to do so. It will expand capabilities and create advancements in data analytics, surveillance, advanced fighting capabilities, and weaponry. The unknowns and the "black box" of AI must be continued to be researched and discussed. Active and open discussions are the best way to weigh the ethical considerations of new AI technology. Transparently approaching the research and development of AI will help safeguard and ensure a positive and transformative impact on national security.

  References

Allen, G. C., & Chan, T. (2017). Artificial intelligence and national security. Belfer Center for Science and International Affairs. https://www.belfercenter.org/sites/default/files/files/publication/AI%20NatSec%20-%20final.pdf  

Amodei, D., Olah, C., Steinhardt, J., Christiano, P., Schulman, J., & Mané, D. (2016). Concrete Problems in AI Safety. arXiv preprint. https://arxiv.org/abs/1606.06565

Bostrom, N., & Yudkowsky, E. (2014). The ethics of artificial intelligence. Cambridge Handbook of Artificial Intelligence, 1, 316-334. https://nickbostrom.com/ethics/artificial-intelligence.pdf

Brundage, M., Avin, S., Clark, J., Toner, H., Eckersley, P., Garfinkel, B., . . . Anderson, H. (2018). The malicious use of artificial intelligence: Forecasting, prevention, and mitigation. arXiv preprint. https://arxiv.org/abs/1802.07228 

Scharre, P. & Lamberth, M. (2022). Artificial Intelligence and Arms Control https://www.cnas.org/publications/reports/artificial-intelligence-and-arms-control 

Taddeo, M., & Floridi, L. (2018). Regulate artificial intelligence to avert cyber arms race. Nature, 556(7701), 296-298. Retrieved from https://www.nature.com/articles/d41586-018-04602-6

Work, R. (2021). Principles for the Combat Employment of Weapon Systems with Autonomous Functionalities https://www.cnas.org/publications/reports/proposed-dod-principles-for-the-combat-employment-of-weapon-systems-with-autonomous-functionalities