Over the past few decades, Western forces have gained the upper hand over adversaries with the help of technological advancements. Tactical ground fighting armies have been driven to win battles due to their technological dominance. This dominance depends heavily on the evolution of the technologies and concepts that are associated with the Information Age. It is expected that the constant evolution of information technology will continue to play an imperative role in shaping warfighting technologies between now and 2050. In this article, I will highlight the impact of Artificial Intelligence (AI) as a transformative technology on the fundamental nature of army operations over the next thirty years.
With technology evolving exponentially, National Defence will continue to research and adopt the latest technology to remain ahead of its adversaries on the battlefield. To remain a contender, the Canadian Armed Forces (CAF) must adopt these technologies as they become mainstream. This is inherently a part of the progress of industrial revolutions. As we continue to embark on the fourth industrial revolution – Industry 4.0 – we will see technologies like cloud computing, Big Data, the Internet of Things, and AI evolving. Of all the emerging technologies that fall under Industry 4.0, one can argue that the true fourth industrial revolution will indeed be AI.
Our ability to comprehend the future and to appreciate its impact on the way defence is conducted is critical for the success of National Defence. It is difficult, however, to truly imagine how the next thirty years will unfold, in particular on the topic of AI. By way of comparison, picture a soldier falling asleep in 1884 only to be awakened in 1916 in the middle of World War I, he would witness technological advances such as the machine gun, long-range indirect fire artillery, airplanes, field phones and the early beginnings of tanks. Although certain concepts and elements of these technological advances existed in 1881, the achievements of technology observed by the time-traveling soldier will be unimaginable. In the same manner, I will attempt to capture elements that partially exist today, and which will completely change the battlefield in 2050.
The Tactical Ground Battlefield of 2050
In 2015, the US Army Research Laboratory assembled experts, intellectuals, and leaders from the public, private and academic sectors to envision the future of the tactical ground battlefield in 2050. The group identified some future capabilities that will influence the battlefield of the future and will undoubtedly play a pivotal function in the future of AI and the Canadian Army (CA).
- Augmented Humans
The augmented human soldier will be a soldier that is “physically and mentally augmented with enhanced capabilities that improve their ability to sense their environment, make sense of their environment, and interact with one another, as well as with unenhanced humans, automated processes, and machines of various kinds.” This soldier will be augmented with cognitive skills, acute vision, and auditory skills, physical enhancement with the help of exoskeletons and computer-assisted decision-making processes. In essence, the augmented human soldier will be smarter, stronger, and able to integrate himself into a network of systems. Due to the advanced abilities of augmented human soldiers, the need for traditional soldiers will be less in the battlefield of the future.
- Automated Decision Making and Autonomous Processes
The decision-making process that commanders make daily during a battle is critical to the success of a mission. The battlefield of the future will be significantly automated allowing for decision agents to be integrated into Command and Control (C2), the Operational Planning Process (OPP), the Intelligence Preparation of the Battlespace (IPB), the understanding of Intelligence, Surveillance and Reconnaissance (ISR), and Battle Damage Assessment (BDA). These agents will conduct tasks such as “filtering information, fact-checking, fusion, dynamic access control (determining who has access to what information), and adaptive information dissemination (who should receive specific pieces of information and/or notifications).” The battlefield of the future will have many sensors deployed for collecting information. The simplest forms are micro-aerial sensors that collect information and remotely send them back to headquarters. Humans will not be able to process, understand, analyse and prioritise the inflow of data from thousands of sensors, as such, automated agents will be capable of understanding the commander’s mission intent and thereby task-specific sensors to deliver mission-specific information. These agents will then sort through the information and discard useless intelligence while retaining what’s most relevant and pertinent.
- Misinformation as a Weapon
With the abundance of information from sensors, operators will work in an information-rich environment. This may cause an information overload in certain cases making it difficult for both humans and AI to confirm the quality of the information – information correctness, authenticity, and security – which will be dependent on the reliability of the source. Similarly, misinformation can be used as a weapon. In an environment of information overload, it will be trivial to deceive the adversary by incorporating believable material that would be misleading and thereby undermining legitimate trust, causing confusion and delaying decisions.
- Micro-targeting and Weapons Miniaturization
The concept of micro-targeting is the “identification and surgical engagement of specific individuals employing either kinetic or non-kinetic means.” In today’s world, it is the equivalent of tailored news feeds or ads on a user’s browser-based on their historical browsing metadata. In the battlefield of the future, micro-targeting will evolve to include the ability to hack the enemy in cyber environments and mine information relating to their precise location. This, coupled with weapons miniaturization, will enable for precise engagement of the enemy resulting in less collateral damage.
- Large-scale Self-organization
Large-scale self-organization will occur as a new means of command and control. The battlefield of 2050 will see “individuals, teams, and software agents, . . . self-organize, dynamically creating, and modifying collaborative processes.” Today, all users of smartphones experience small-scale self-organization where applications self-synchronize, collaborate with other applications, and even dynamically adapt to the user’s habits. The battlefield of 2050 will see self-organized and autonomous entities dramatically evolving their behaviours in response to the changing environment. This truly redefines the network-enabled C2 concept and sets new limits to agility in the battlefield. Large-scale self-organization will be made possible with the help of Deep Learning (a branch of AI) where agents will be able to learn algorithms and sets of architecture enabling them to automate the extraction of “complex representations without human intervention.” Today, several commercially available agents already demonstrate deep learning capability. Google’s AlphaGo machine had a historic victory over a human in the ancient board game Go, which demonstrates that a machine can have approximated human intuition and can outsmart the best human competitor in the game. With the help of deep learning, AI agents will continuously learn, adapt, and overcome their adversaries in a networked fashion to win battles.
The Impact on Fundamentals of Army Operations
In Waypoint 2018 – The Canadian Army Advancing Toward Land Operations 2021, LGen Hainse states that “…in the digital age, network-enabled forces that can see, understand, and act more effectively in their operating environment will have a significant advantage over their adversaries.” This document will help the “Army leadership in making decisions for the modernization of the Canadian Army” as part of its transformation to realize the force employment concept articulated in Land Operations 2021: Adaptive Dispersed Operations: A Force Employment Concept for Canada’s Army of Tomorrow. The fundamentals of army operations listed in Land Operations 2021 and Waypoint 2018 were written purposefully with the broad language to “meet an unpredictable future”. The following section will highlight the details given the predictable future explained previously.
- Adaptive Dispersed Operations
The future battlefield will continue to be based on the conceptual designs of Adaptive Dispersed Operations (ADO) requiring Land Force intervention in the full spectrum of threats. Nevertheless, the balance of that spectrum will shift from the physical plane of today to the moral and informational plane. This will be primarily due to a significant shift in the battlespace. Although battles will continue in air, land, and sea spaces, it will be expected that, due to Industry 4.0, the battlespace balance will lean more towards the electromagnetic spectrum and the information environment. With the help of micro-targeting and large-scale self-organization, AI agents will be able to set favourable conditions before even contact with the enemy.
The fundamentals of dispersed operations of the find, fix, and the strike will be much easier to achieve in the battlefield of the future. AI sensors of all sizes and shapes with the ability to comprehend the mission’s intent will be able to task targeted sensors allowing for a precise and expedient find. Once again, the battlespace will not necessarily be limited to land, air, and sea, but also the cyber domain. For example, an AI agent may place a digital sensor on e-mail traffic, continuously seeking key information that may potentially link to the enemy. Micro-targeting will thus be achievable with precise fixing and striking of the enemy, before contact in the physical plane.
A more important impact on the fundamentals of Army operations will be the commander’s significant increase in Shared Awareness (SA). With the help of autonomous analysis of the sensory information, AI will be capable of implementing decentralized decision making autonomously and instantly deliver the most relevant and precise SA to the command staff. Commanders will be capable of making assisted decisions to continue winning in the battlespace. ADO will be managed with the help of AI systems and agents that can mitigate the dispersion of the enemy with the help of its sensors, while delivering detailed fix and strike effects. Notwithstanding, the human command will be critical for the success of the mission through SA, as highlighted in Waypoint 2018, “The centrepiece of ADO remains the Canadian soldier, but the key enabler is its network.”
- Operational Functions
The five operational functions will continue to form the fundamental framework of Army operations. Moreover, operational functions Sense and Act will evolve to be the strongest functions enabled by AI and autonomous agents. As AI becomes well integrated into our force structure and gains deep learning capabilities and the ability to conduct large-scale self-organization, human soldiers will have a very minor role to play in the sense and act functions.
The Shield function will be shared between humans and AI agents in the physical and moral plane, leaving autonomous machines to provide the necessary protection for its survivability and freedom of action. In the information environment, cyber defence will play a pivoting role in the protection of AI systems in its most vulnerable plane.
As fewer human soldiers will exist on the battlefield, the Sustain function will evolve to exist in the information plane significantly more than the physical and moral plane. The provision of materiel and personnel support will become less of a requirement. Sustainment of AI agents on the other hand will require specific expertise that may be beyond the future soldier’s technical competency. As such, soldiers will either be required to gain technical robotics, programming, and mechanical skills or DND will see the dependence on contractors and commercial services for the maintenance, training, and re-generation of AI systems. Additionally, the requirement for first-line support may no longer exist as AI systems will be capable of self-diagnosing and self-healing.
Finally, the centre of gravity of the future battlefield will remain the Command function. It will be here that the human commanders’ role will be critical in shaping mission command and integrating all the other functions into the strategic, operational, and tactical levels of the battle. After all, command is the creative expression of the human will necessary to accomplish a mission, and although the commander will be enabled in the decision-making process, it is the human will that will help in achieving the end state.
- The Human Dimension
An important question looms in the discussion of AI in defence: Will there be a requirement for humans in the battlefield of the future? Contrary to Hollywood and science fiction depiction, most certainly yes. Humans will continue to play a critical role in the battlefield of tomorrow especially in humanitarian and peacekeeping operations. That said, full military kinetic operations will certainly see a significant reduction in the number of soldiers reduced from the force structure. As the cost of soldiers continues to increase, it will be much more cost-effective to employ AI, autonomous, deep learning systems and agents to accomplish missions. The kinetic battlefield of 2050 will see a force structure where the lowest level of the human soldier on the ground would, most likely, be the equivalent of today’s Platoon Commander, directing a set of automated AI-enabled platforms.
The human dimension also plays an important role which AI is unable to learn – such as values, culture, relationships, soldier identity, and “the psychological contract between the soldier and society”. The most important of all is the human ability to deal with ethical dilemmas. How will AI deal with ethical dilemmas when placed in a lose-lose predicament? The famous example is based on science fiction author Isaac Asimov in his classic story Liar where an AI robot struggles to respect two of three laws he was programmed to obey, demonstrating that humans will indeed provide a compassionate and human aspect to the battlefield that would be otherwise irreplaceable.
- Cyber Defence
The fourth industrial revolution will make AI available to the public and the commercial sector. Militaries and governments will not be the only operators in the informational environment and militaries will be equally exposed to the same threats they impose on their adversaries. As mentioned earlier, the battlespace will be predominantly in the electromagnetic spectrum and information environment. This will cause cyber defence to become a vital asset to preserve the integrity of our AI systems and prevent us from being victims of our systems if it turns against us. Equally, misinformation used as a weapon can be detrimental to the decision-making process, causing wrong courses of actions and leading to mission failure. It is obvious that cyber may become the fourth environment in the CAF and cyber defence may become a trade within which programs and AI systems will necessitate the shielding of the information environment.
The Future Battlespace
AI will undoubtedly play a role in the CA of the future in one way or another. At the very least, by 2040, the CA should see initial variations of the Augmented Human. For the most part, if employed correctly, AI can completely change ADO and with the help of its automated decision making, micro-targeting, large scale self-organizing, and its ability to sort through endless data, it will enable the CA superiority in the battlespace. Moreover, humans will continue to play an important role as they possess human attributes that will never be programmed into AI.
This article was originally written as a service paper for the Canadian Forces College and reprinted here by permission. LCol Amir ElMasry, CD is currently serving as Deputy G6 for the Canadian Army.