Proactive Collaboration and Security Vigilance Key to an AI-driven Renaissance in Defense

Blog

Proactive Collaboration and Security Vigilance Key to an AI-driven Renaissance in Defense

New AI Fight Requires Investing in AI & Quantum Technologies, Recognizing the Benefits & Threats the New Tech Represents, while Incentivizing Commercial Innovation

By Joshua Cryer

To meet the threat of a changing battlespace dominated by artificial intelligence (AI), the U.S. defense community must rethink how it collaborates with the commercial sector. Joshua Cryer, a 25-year veteran of military communications and currently president and CEO of Reticulate Micro, argues that today’s AI fight requires new mindsets that reject the status quo, a laser focus on AI and quantum-based technologies, and incentivizing commercial innovation.

https://reticulate.io/wp-content/uploads/2024/06/afcea-banner.jpg

Artificial Intelligence (AI) is revolutionizing many fields, from finance and insurance to automotive and medicine. According to a recent McKinsey Global Institute study, AI adoption overall jumped to 72 percent in 2024, up from 55 percent last year.

The defense market is also seeing an emerging AI-enabled renaissance. AI products dominated SOF Week’s exhibit floor in Tampa in May 2024. The world’s largest special operations community gathering unveiled capabilities that can reduce the cognitive load on warfighters, enhance force protection and augment long-range reconnaissance capabilities. US SOCOM officials also emphasized that the future of AI for special operators will center on “collaborative, portable autonomy,” where unmanned sensors can work together collaboratively and autonomously.

But getting AI capabilities from vision to reality is a complex challenge, with implementing AI in the tactical environment taking significantly longer than advances in the commercial sector.

A key challenge is the Pentagon’s procurement processes and how it handles data management. Last summer’s House Armed Services Committee hearing assessing the barriers inhibiting the Department of Defense (DoD)’s adoption of AI was telling: It found that while the Pentagon generates over 22 terabytes of data daily, outdated data retention and management policies result in data that is not “AI ready,” and therefore, can’t be tapped to its full potential by warfighters, analysts and operators, according to testimony by Alexander Wang, founder and CEO of Scale AI. The data management challenge is complicated further by advances in quantum computing and AI posing significant vulnerabilities to command-and-control systems and databases. Advanced Encryption Standard (AES) 256, considered an NSA Suite B gold standard for U.S. government encryption, to protect data being transferred or stored, is under attack thanks to advances in both AI and quantum computing. Addressing the quantum threat requires new mindsets in how the government innovates with the commercial sector.

Supplanting the Status Quo with New Wave of Innovators & Capabilities

Aerospace and defense primes have long wrestled with the balance of delivering emerging technologies to the military and maintaining a profitable business. New technologies often present a threat to traditional business models that recoup non-recurring engineering (NRE) once a product reaches full rate production (FRP). Margins are typically further enhanced with incremental upgrades to those product baselines over time, enabling the enjoyment of maximum profitability – a joy to shareholders and incentivized execs alike. Opposite these factors is the growing defense need for advanced technologies to counter state actors that have ramped up their ability to deliver sophisticated defense-related technologies. The loser in this scenario is the military end user; the warfighting soldiers, sailors, marines, airmen, and guardians who need the best of what’s available to counter near-peer threats.

Emerging from the ecosystem of slow-moving juggernaut primes are a few forward-leaning and hungry firms like Anduril that challenge this paradigm. Anduril is powered by a war chest of capital: its $1.5 billion funding round last year dominated U.S. defense tech startup raises, according to Crunchbase data. Anduril is unencumbered by existing business franchises and driven by a powerful vision to super-enable our nation’s warfighters with the latest technologies. The firm is decidedly a technology company focused on delivering disruptive technologies to defense – an activity often touted by the primes that, in reality, only deliver incrementally improved legacy technologies.

On the technology front, recent advancements in AI and quantum computing (QC) are helping shift the status quo. The marriage of the two technologies will be profound, but the ultimate outcomes are still unclear. The DoD is aware of the massive potential for enabling both good and evil activities. Young Bang, Principal Deputy Assistant Secretary of the Army (Acquisition, Logistics and Technology,) addressed this concern at the Army’s Digital Transformation Panel at last fall’s AUSA Expo and reported by Breaking Defense. There, he advocated for industry support for establishing a security framework for AI. As the Army pivots to embrace the benefits of AI, there is an understanding that this still-nascent technology will realize the unprecedented analytical capabilities dreamed of by commanders past while simultaneously presenting harrowing security concerns that are just as disruptive to current security paradigms as the technology itself. The benefits are worth the risk and certainly merit the cautious approach to secure implementation.

AI & Video Driving Process Improvements & Intelligence at the Tactical Edge

Advancements in discrete sensing technologies, concentrated edge-computational technologies, and multiplying drone swarms have exponentially increased video and sensing information products, and a growing multitude of connection options have turned the fog of war into a cloud of confusion. The information is far too much for individuals or groups of individuals to consume, normalize, and action in any reasonable manner. The hope is AI will provide an enabling capability that sorts through the minutiae in the cloud to provide meaningful and actionable intelligence, helping our warfighters execute their functions with greater effectivity and safety. The technology has massive potential for video and image analysis, connectivity management, and supporting decision-making processes.

AI can “watch” video and sort through images faster than a human being or group of human beings ever could, identifying suspects in the shadows, the gait of a bad guy in a large crowd, or that hostage through facial or voice recognition. So long as it is powered by the right number of computational resources and connectivity, AI can do this while people sleep, operating 24/7. AI will become a fundamental ingredient for all future Processing, Exploitation, and Dissemination (PED) activities.

PED is essential to the Army Intelligence Process that transforms raw video and data into usable information for additional analysis. PED outputs are typically used by combatant commanders and their supporting staff to inform the commander’s situational understanding, helping them make better decisions. AI brings a transformation automation potential to the PED cycle of planning and directing, information collection, producing intelligence, and then disseminating it to command and staff channels. Guided by commanders’ intent, AI will enhance every aspect of the intelligence process cycle, helping to focus collection planning to avoid “dry holes” where sources end up not providing important information, boost collection approaches with AI-enabled efficiency, expedite production activities that make sense of collected information, and ensure intelligence information gets to the right folks at the right time. AI will revolutionize the entire intelligence apparatus of the military.

AI will also improve the operational activities at the tactical edge, ensuring our warfighters keep and maintain the unfair advantage at the frontline. Future combat vehicles will use AI to navigate the battlefield with ease, automatically detecting and avoiding obstacles, determining the fastest or most tactically advantageous routes, or identifying hostile targets for expedited prosecution. Radios and satellite terminals of the future will combine AI with advanced discrete sensing technologies to detect interference and congestion in the operational environment, and in response, automate waveform, frequency and power configurations to counter enemy jamming or avoid detection.

Security is an essential aspect for AI implementations. The future of this technology field has massive positive and negative potential. As access to AI and quantum computing grows and the technology matures at an exponentially rapid pace, state actors and peer threats may be the least of our military’s concerns. It won’t be long until eager and bored 14-year-olds, fueled by energy drinks and gummy worms, can access cloud-hosted quantum computing and AI clusters to hack into bank accounts, Instagram accounts, and government computers hosting national security secrets.

Thanks to advances in quantum computing, current encryption security methods, practices and algorithms are “dead men walking.” For example, encryption based on Rivest-Shamir-Adleman (RSA) and Elliptic Curve Cryptography (ECC)approaches are effectively compromised by quantum computing. QC can use Shor’s algorithm to factorize large integers, which is the basis of RSA. The approach uses quantum-based mathematics to factorize large numbers in polynomial time, effectively breaking RSA encryption. The same approach can swiftly solve the complex math underlying ECC, as QC can eavesdrop, and decrypt ECC-based key exchanges in near real time.

Embracing Quantum-accelerated AI

Coupling AI with quantum computing will affect cybersecurity in both the commercial and defense sectors, significantly complicating the security landscape, particularly concerning cryptographic systems. There are several ways that this powerful combination will likely impact cybersecurity, both empowering cybersecurity technologies and hyper-enabling nefarious activities.

Quantum-accelerated AI algorithms, particularly machine learning (ML) models, will be able to process vast amounts of cryptographic data at an accelerated pace, far more efficiently than classical computers. This will lead to more effective cryptanalysis that will break classical encryption schemes at an exponential speed. Weaknesses in classical cryptographic algorithms will be discovered more quickly, and the combination will enable the prediction of keys through sophisticated pattern recognition. This means extant security systems that protect our bank accounts, medical files, conversations with legal counsel, and private pictures on the cloud will be broken overnight. The potential for the “gray rhinoceros,” referring to a highly likely yet ignored threat, first coined by Michele Wucker in her book, The Gray Rhino: How to Recognize and Act on the Obvious Dangers We Ignore, to turn into a black swan event increases moment by moment.

AI-based systems are already used today for detecting and responding to internal and external cyber threats. The systems analyze patterns of behavior and identify potentially threatening anomalies, helping cybersecurity professionals counter hackers and automated malicious logic. Coupling QC to this AI-powered technology will exponentially increase the efficiency and speed of these analyses. While there is massive potential for using this to enhance cybersecurity systems more effectively detect and counter threats in real time, conversely, bad actors will use this same technology combination to develop more sophisticated, evasive, and resilient malware that is difficult to detect.

Quantum computing will further enhance machine learning models to create the field of Quantum Machine Learning (QML). QML will make machine learning models more accurate and capable of processing infinitely more complex datasets. These improved models will lead to more sophisticated AI systems that can launch cyber-attacks and defend against them. For example, AI-driven intrusion detection systems (IDSs) will become far more accurate, identifying complex multi-vector orchestration attacks. At the same time, attackers will use this same technology to develop AI-empowered attacks systems that will better mimic legitimate user behaviors, making detection more difficult.

The potential negative impact of these technologies on military operations is massive by adversaries equipped with AI and quantum computing-enabled technologies. When peer and near-peer hostile forces implement these technologies in the battlefield, they will be able to decrypt communications that rely on classical cryptography in near real time. When applied to communications, the potential is vast for these dual technologies to analyze and counter radio frequency waveforms that are currently being used to hide signaling with classified mixes of obfuscation approaches, enabling the direction finding of communicating technologies. Further, AI+QC technologies could also be used to observe the movements of maneuver elements, and better predict their next steps, giving hostile forces the upper hand in real-time mission planning.

Call to Action: Embracing the Long Game & Incentivizing the Industry to Innovate

In conclusion, to best realize the positive effects of Artificial Intelligence, and the further coupling with quantum computing, we must take a proactive stance on rapidly adopting these emerging technologies while enhancing our vigilance to protecting the advancements. As Mr. Bang pointed out, defense, academia, and industry must work to effectively collaborate in a proactive manner. While these technologies can significantly enhance our security measures, intelligence apparatus, and communications technologies, they also pose new, novel, and unforeseeable risks that require proactive and strategic action from our collective community.

Defense firms must look beyond the quarter, understanding that the AI fight is both a fundamental and financial long game. While there have been acquisition improvements in the form of other transaction authorities (OTAs), the DoD must continue to improve these processes to enable access to and implementation of emerging technologies – and create structures to incentivize industry to innovate. Continued investment in research and development of advanced AI and quantum-resistant cryptographic algorithms is essential for mitigating these emerging threats.

About the Author

Joshua Cryer is president and CEO of Reticulate Micro, a mission-focused defense tech company redefining resilient communications. A U.S. Air Force veteran with 25+ years of military communications experience, Joshua has overseen communications for the elite United States Special Operations community and NATO as well as for the defense contractor community. He leads Reticulate strategy to deliver tactical satcom, terminal management and secure video streaming solutions in places that have never had video before. Through its partner Quantropi, Reticulate is building quantum-level security and trust across its product offerings.

Copyright © 2024 Reticulate Micro, Inc. | All Rights Reserved.