Algorithmic warfare is a warfare which is driven by autonomous systems powered with Artificial Intelligence (AI) that can control the weapon platforms, systems, operations and even tactics in a network centric battlefield. Algorithms are sequence of instructions and rules that computer machines use to solve problems. Algorithms can be as simple as a decision tree or a complex set of instructions governing the autonomy of weapon systems.
Human lives are becoming more valuable over the years. Induction of unmanned autonomous systems into the battle field could eventually improve the speed and accuracy of weapon systems as well as bring down the battle field causalities. Major military powers are currently planning to switch to algorithmic warfare making combat actions faster than the decision-making speed of humans. A better algorithm will ensure technical superiority over adversaries in processing, analysing and rendering solutions.
In an Algorithmic Warfare, the solutions provided by algorithms would shape the military judgments, tactics and decisions. In short, the ‘Intelligent Machines’ or ‘Killer Robots’ will fight the war for their respective masters. We are now entering into a new era of RMA (Revolution of Military Affairs) and MTR (Military Technical Revolution).
Artificial Intelligence and Algorithmic Warfare
Artificial Intelligence (AI) is a man-made intelligent system with skills such as learning, reasoning and solving problems. In Algorithmic Warfare, both weapon platforms and military assets fielded by the fighting force would form a part of the AI based battlefield network. The specific task of weapon systems and associated platforms engaged in war would be designated by complex algorithms in the time and space domain.
There is always an element of surprise for enemy by virtue of the speed and accuracy at which smart weapons are operated. Intelligent machines are capable of self-learning and rendering optimised output suiting to the tactical situation. Machine learning (ML) is inherent in AI and the system with integrated neural network would learn continuously and evolve during operations.
There are many ethical, moral and legal issues associated with algorithmic warfare which is driven by AI and Machine Intelligence (MI). Most of the autonomous weapons fall under the ambit of ‘Lethal Autonomous Weapon Systems’ (LAWS). Many nations have rallied against LAWS at the United Nation. Efforts are on by human right activist groups at UN to institute a complete ban on development of autonomous weapon systems. Nevertheless, leading military powers continue their research and development on AI based weapon systems.
Military operations are lethal and lead to loss of human lives. Availability of combat manpower for military operations would decline in the coming years. Since the cost of man power is increasing year after year, advanced nations are planning to do away with the expensive conventional war preparation and switch to Algorithmic Warfare. The cost of maintenance of defence personnel is going up every year. Unmanned autonomous systems could be a great option available for nations to reduce the overheads and minimise battle field casualities.
Autonomy to Sub-systems
During a combat, military command would engage with many weapons and sensor systems simultaneously. It is humanly impossible for the command to control and monitor supplementary systems along with core weapon/tactical system. Hence forth by rendering autonomy to sub-systems would help the command/crew to focus on the core operation leaving the peripheral systems under the control AI.
Disruption in Conventional War Fighting
AI powered autonomous systems can change the conventional way of fighting war and may change some established precepts in future. The new approach would even force the military command to re-structure the fighting force. The size of an armed force may become irrelevant compared to the battle being fought. Wealthy small states could field more potent forces than the large populous nations. Intelligent machines may reorient the forces and minimise the advantages of some states with respect to their experience and skill. Majority of the unmanned autonomous weapon systems deployed can be expendable. More miniaturised stealthy weapon platforms will come into fray in future.
Currently there are many AI based autonomous systems in the field of navigation, tactics, self-protection, weapon guidance and communications. AI is being introduced into many areas including cyber-security, logistics/ transportation, target recognition, health care, simulation, training, threat monitoring, tactical situation analysis, prediction and data analytics.
Engaging enemy drones and aircraft, integration of tank guns with sensors, coordinating resupply and maintenance of military assets, planning and installation artillery barrages, fusion of different sensor feeds together into a coherent picture and analysing/identifying blind spots in enemy defence are few applications where in AI has delivered positive results.
We can broadly classify autonomous systems into two categories. Autonomy achieved at rest and the autonomy while in motion. The one at rest is superior and virtual which would facilitate and control the entire battle order at a higher level. The one in motion refers to the autonomy of moving platforms and weapons. Both categories combine to form a coherent autonomous entity while engaging in a war.
Network Centric Algorithmic Warfare
Network centric warfare has improved the coordination between fighting units and speeded up the data exchange between platforms and command. Protocols were laid for seamless communication between heterogeneous systems and interoperability of divergent platforms for utmost precision and speed. Networks are excellent in collecting, storing and sharing information but less useful in processing and delivering useful insights to command.
Networking of autonomous platforms is the next level of integration; wherein multi-dimensional divergent platforms synergise to form a combined force in a battle field. Visual data capture and analysis are key for decision making and successful missions. Dedicated secured military cloud and standalone captive battle field network mandatory for cyber-secure network centric operations.
Integration of Machines and Humans
Both humans and machines have strengths and shortcomings. The notion that machines could replace humans is future is not completely true. In algorithmic warfare, the real implementation involves deployment of an optimum mix of human and machine intelligence together. Battle networks comprise interlinked computer systems programmed for data analytics, information gathering, sensing hostile situations and rendering command which overlays the theatre. AI based systems will be supplemented with Deep Learning neural networks.
AI enabled network thus facilitate weapon systems complemented by a large number of low cost autonomous unmanned systems designed for specific tasks. The unmanned systems are designed and fielded for unique tasks and are expendable, hence could be risked for dangerous missions in lieu of manned platforms.
The main challenge in deploying AI based systems is the lack of reliable data for drawing meaningful conclusion. Collection of useful data for analytics may take many years. Integration, testing and deploying of autonomous weapons safely is a big challenge. Initial cost involved in development of systems and generating meaningful solutions will be very high. Only few manufactures in the world can provide military standard hardware support to meet the requirements. Identifying suitable development teams for AI platforms is another challenge which cannot be undermined.
AI based systems are not flexible and are incapable of handling minor changes in context. AI systems have poor domain adaptability and hence they struggle to apply knowledge learned in one context to another. Humans generally make better judgments in situations of high complexity and uncertainty. This means that amalgamating men and machine would be optimal and can become more potent.
Advancement of AI is irreversible and Algorithmic Warfare will become a disruption to the existing conventional warfare. Application of Artificial Intelligence and Machine Intelligence has increased manifold in the recent years for a wide range of military applications. Algorithmic Warfare could trigger seamless integration of military assets and associated tasks. This warfare can also impact the design and development of weapon platforms in future. Russian President Vladimir Putin recently declared that ‘whoever becomes the leader in the AI sphere, will become the ruler of the world’.
- Algorithmic Warfare Applying Artificial Intelligence to Warfighting by LAYTON PETER (2018).
- The U.S. military, algorithmic warfare, and big tech by KHARI JOHNSON (2019)