The terrorist attacks on September 11, 2001, have been described in myriad ways but, above all, they were a watershed moment in U.S. history that fundamentally reshaped society, politics, the defense mindset, and the national security apparatus. An inarguably significant effect of that tragic day has been the rapid acceleration of investment in artificial intelligence (AI) and machine learning (ML) technologies for national security purposes. As a response to the events of September 11th, the U.S. intelligence community (IC) and the Department of Defense (DoD) initiated a paradigm shift toward more agile, data-driven approaches to identifying threats and mitigating risks. In the two decades that have followed, AI and ML have moved from peripheral tools to central components in the strategy and tactics of national defense and intelligence operations. In this paper, we discuss some of the more transformational technologies.
Autonomous Systems and Robotics
Unmanned aerial vehicles (UAVs), including drones of every description and function, have become ubiquitous in intelligence, surveillance, and reconnaissance (ISR) operations. They rapidly became critical components of armed conflicts, as well. Neither the U.S. Air Force nor the Central Intelligence Agency (CIA) had an armed drone in 2002, but drone strikes had become increasingly important by the end of 2001.
The U.S. military’s drone arsenal now ranges from the large, powerful, heavily armed Predator 2, or MQ-9 Reaper, to the silent, precise, tube-launched SwitchBlade, to the hand-launched, four-pound Raven, or RQ-11. AI-driven autonomous systems reduce risks to the human warfighters and expand operational capabilities, including enabling over-the-horizon (OTH) strategies. Algorithms capable of making real-time decisions based on sensor data supplement, and in some cases are replacing, human operators, which enables extended mission duration and reduced cognitive load on personnel.
Enhanced Data Analytics
Early investments following the attacks on September 11th focused on data aggregation and analytics, in which AI systems were primarily used for sorting and sifting through vast amounts of data. Subsequent advances in natural language processing (NLP) and predictive modeling have allowed the daily, multi-terabyte congeries of unstructured data to be converted into actionable insights. Today, intelligence analysts leverage ML algorithms and generative AI (GenAI) models to analyze data from video, audio, and textual communications, imagery from space-based, air-borne, terrestrial, maritime, and submarine sensors, and metadata from myriad other sources, thereby increasing the efficiency of threat detection operations.
Cybersecurity
As the world became increasingly interconnected, cyber threats emerged as a distinct challenge to national security, as well as to civilian and military infrastructure, financial institutions and other corporate entities, academia, social networks, media entities, and governments at every level. Every standard partition of concern—hardware, software, the cloud—is vulnerable. Even information in transit is not immune from threats such as signal/telemetry interception, deliberate signal attenuation, and traffic diversion. The attackers include state actors, such as hacking “teams” funded by Russia, China, Iran, and North Korea; non-state actors, such as hacktivists and terrorists; criminal organizations of every stripe; and talented amateurs.
AI and ML algorithms are being deployed for intrusion detection, anomaly spotting, and immediate threat neutralization, acting as force multipliers for human cybersecurity teams and fast becoming indispensable for maintaining the integrity of defense networks and critical physical, economic, and social infrastructure. The U.S. National Security Commission on Artificial Intelligence (NSCAI) wrote presciently in its final report in 2021 that adversaries understand that ML is a powerful tool for gathering and analyzing data, and that AI can be used to pinpoint and exploit vulnerabilities in individuals, society, and entities of every size and description.
A 2021 study showed that automated adversarial or intelligence-gathering bots linked to foreign hackers, as well as foreign intelligence agencies, interrogated U.S. Internet protocol (IP) addresses seeking entry to vulnerable industrial controls systems. The bots used various obfuscation techniques to disguise their origin and deployed wiper and scraper technologies to easily gather copious data.
Human-Machine Teaming
AI is also revolutionizing the human element in national security through decision-augmentation capabilities. AI-driven situational awareness tools are becoming standard components of command and control systems. By providing high-quality, real-time information devoid of “noise” in the data, these systems enable human analysts and decision-makers to focus on the most relevant information and use their expertise and context-awareness to interpret the data, assess the situation, determine the appropriate response, and take the appropriate action in high-stakes, time-sensitive environments.
Ethical Considerations
As AI and ML technologies advance, there is growing recognition of the need to establish a rigorous ethical framework with clear guidelines, periodic independent reviews, and oversight mechanisms, including human-in-the-loop (HITL) or human-on-the-loop (HOTL), for critical functions. Concerns that must be addressed range from questions of accountability and fairness to potential unintended consequences or escalation and necessitate a rigorous ethical framework, particularly when life-or-death decisions are automated or augmented by AI.
In the 20-plus years since the attacks on September 11, 2001, AI and ML have evolved from experimental technologies to essential components of the U.S. national security strategy. They are not just tools; they are force multipliers that enhance the effectiveness and reach of human operators and decision-makers. However, as we move forward, striking the right balance between technological innovation and ethical responsibility remains critical. The continued stewardship of leaders across the military, IC, industry, and academic sectors will be crucial in shaping the future role of AI in safeguarding our nation, enabling us to honor the lessons learned from the past and equip ourselves for the complex security challenges of the future.
