Regulation of Artificial Intelligence in the Military Sphere: Limitation or Encouragement?

It would be surprising if the technological capabilities of AI were not applied in the military sphere. According to the influential British publication The Economist, AI is changing every aspect of warfare. Unfortunately, today Ukraine itself is forced to check this statement. 

Researchers of arms markets where AI is used name ten types of its implementation for military purposes: autonomous weapons systems, surveillance and reconnaissance, recognition and classification of targets, analysis and prediction of threats, logistics and management of supply chains, cyber security protection. In addition, there are also such types as electronic warfare, simulation and training, medical diagnostics and treatment, decision support systems.

The problem of applying AI in the military sphere prompts states to make joint decisions regarding the regulation of the norms of "soft" international law. The Responsible Artificial Intelligence in the Military (REAIM 2023) summit, held on February 15 and 16, 2023 in The Hague, adopted a Political Declaration on the Responsible Use of AI and Autonomy for Military Purposes. This declaration is aimed at forming an international consensus on responsible behavior and state management of the development, deployment and use of military AI. As of February 12, the declaration was supported by 54 states, including the Group of Seven (G7) countries and Ukraine.

We will continue the research of the experience of the leading countries in the application of AI, which we started earlier.

 

The United States of America

The United States Department of Defense's first artificial intelligence strategy was developed in 2018 and published in February 2019 (updated in 2020). The strategy covered such topics as strategic goals, strategic approach and strategic directions, and was focused on increasing the efficiency of using modern AI capabilities.

The United States Department of Defense did not ignore the training of the military, having prepared the Strategy for the training of military personnel in the relevant areas of AI. Its main goal is "accelerated implementation of artificial intelligence capabilities at scale."

And already on February 2, 2022, the Ministry of Defense of the United States of America published the Software Modernization Strategy. The motive for its development was quite simple: "To create a stronger lethal force, we need the ability to develop faster and be more adaptive than our adversaries."

In March 2023, the United States of America adopted a National Cyber Security Strategy. This is what was said in the text of this strategy about the reasons for its adoption: "The governments of China, Russia, Iran, North Korea and other autocratic states with revisionist intentions are aggressively using advanced cyber capabilities to achieve goals that are contrary to our interests and generally accepted international norms."

On February 21, 2024, U.S. Deputy Secretary of Defense Kathleen Hicks presented a presentation on "The Global AI Rivalry" at the 2024 Defense Data and Artificial Intelligence Symposium in Washington, DC.

The report names the most important factors prompting the Ministry of Defense of the United States of America to pay extraordinary attention to the development of AI. The first factor is the need to use this technology, which provides "an even greater decision-making advantage than the United States has today, given the challenge we face from the People's Republic of China."

The second factor is the scale of management. Kathleen Hicks compared the United States Department of Defense to "the world's largest enterprise," which has "nearly 3 million people on the payroll; a health care system that serves more than 9 million military personnel, retirees and family members; and worldwide assets spread across 25 million acres, roughly the size of the state of Kentucky. The value of the assets of the United States Department of Defense worldwide is greater than that of Amazon, Apple, Microsoft and Walmart combined.” That's why we need to use data and AI to be "smarter, faster and better stewards of taxpayers' money."

Four different digital technologies, artificial intelligence and analytics groups were merged into the United States Department of Defense as early as 2022 under the leadership of the Chief Digital and Artificial Intelligence Office (CDAO), responsible for accelerating the implementation of data, analytics and AI in the United States Department of Defense. The agency was created through the merger of the Joint Artificial Intelligence Center (JAIC) and the former Defense Digital Services (DDS).

The key mission of the Chief Digital and Artificial Intelligence Office (CDAO) is to regulate and scale the use cases of AI-enabled tools, including executive analytics to measure progress on national defense strategy, the overall operational picture for combatant commanders from operational to strategic levels within the Advancing Data and AI (ADA) initiative, and best-in-class tools and analytics to assist senior leadership and combatant commanders in a dynamic campaign.

On November 14, 2023, the representatives of the Chief Digital and Artificial Intelligence Office (CDAO) released the Responsible Artificial Intelligence (RAI) Toolkit, which is a major output of the United States Department of Defense RAI S&I Pathway signed in June 2022.

The RAI S&I Pathway describes as many as 64 areas of effort to implement ethical AI principles within the United States Department of Defense, including the development of technical tools and guidance to help its personnel design, develop, deploy, and responsibly use AI systems.

 

The United Kingdom

In the United Kingdom, as in all other cases, public decision-making regarding the military use of AI takes place in accordance with the principle of good governance. The system of strategic and regulatory documents is built according to a strict hierarchy and based on the results of consultations with representative bodies of state power, experts who have a real successful long-term experience of implementing projects in the relevant field. Document preparation procedures are carried out as transparently as possible for society.

In particular, within the framework of the national strategic order and strategies, one "comprehensive" level document was adopted, two – on digitalization issues, and one – on AI. The Ministry of Defense of the United Kingdom approved four documents of the "comprehensive" level, nine related to digitization and one related to AI. And finally, the British Army Command approved two "comprehensive" level documents, one on digitalization and one on AI.

So, in October 2023, taking into account just such a structure, the British Army published a strategic document entitled "The British Army's Approach to Artificial Intelligence. A Guide to Accelerating the Army's AI Implementation and Army AI Training," which outlined how the Army plans to become "AI ready" as early as April 2024.

This document accompanies the September 2020 Integrated Operational Concept, the 2021 Defense Digital Strategy and the June 2022 Defense Data Strategy, Defense Artificial Intelligence Strategy. In addition, this document establishes the practical implementation of the Army Digital Information and Data Plan 2023-2025, published in April 2023. In addition to skills, ethics, and internal leadership and data management, the strategy covers the Army's involvement in the private sector.

To implement AI programs in the defense sector of Great Britain in 2021, the Center for Artificial Intelligence of the Ministry of Defense (MOD) was created, which works in close connection with three other units. Among them stands out the Defense Digital organization, which is responsible for the transfer of effective digital and information technologies to the military. Defense Digital has an annual budget of more than £2 billion and a team of around 2,400 staff, including military, civil servants and contractors, and is driven by the Defense Digital Strategy. The unit, called Defense Equipment and Support (DE&S), was created as an independent body of the Ministry of Defense in 2014 with unique management powers delegated by Her Majesty's Treasury and the Cabinet of Ministers Office. The unit operates in the context of four main strategic challenges: Russia's invasion of Ukraine, the rise of China's economic, military and technological power, digital warfare and climate change. The Defense Science and Technology Laboratory (Dstl) is an executive agency of the Ministry of Defense that provides world-class expertise and delivers advanced science and technology for the benefit of the nation and allies.

And one more fresh news that may interest some people's deputies in the Verkhovna Rada of Ukraine. On December 14, 2023, the Defense Committee of the British Parliament created a subcommittee on AI in the field of defense, which, in particular, takes care of the following issues:

1. How clearly has the representatives of the Ministry of Defence (MOD) prioritized the AI capabilities and expertise it believes the defense sector of the United Kingdom needs to have, what are the key drivers it has identified and are they achievable?
2. What strengths and experience does the industry of the United Kingdom currently have in AI in the context of defense applications?
3. How can the Government of United Kingdom best develop capacity and expertise in domestic industry in sectors such as engineering and software to support the development and delivery of defense AI applications?
4. What can the government do to help integrate AI companies of the United Kingdom into defense supply chains both at home and abroad?

The Parliament of the United Kingdom has already had time to thoroughly investigate some of these issues.

 

China

Both the United States of America and the United Kingdom consider China's growing economic, political, and military influence to be the most threatening trend for global security. This also applies to the development of China's capabilities in the application of AI in the military sphere.

Fact: China has participated in summits and signed two non-binding legal documents adopted at major international summits: Political Declaration on the Responsible Use of Artificial Intelligence and Autonomy for Military Purposes (Netherlands, 2023). In addition, China's political leaders signed the Bletchley Declaration agreed at the AI Security Summit in Great Britain (2023).

According to Institute for the Study of War expert Matthew McInnis in the study "Russia and China: A Look at the Future of War": "China's military modernization efforts are aimed at achieving decision-making dominance through a three-pronged approach. It includes such aspects as doctrinal transformation and ideological rigidity; the use of advanced technologies to shape the nature of modern conflicts; innovative training methods to compensate for the lack of combat experience during the war."

In addition, the available information about contacts at the highest level between China and Russia as early as February 2024 regarding cooperation in the development of AI in the military sphere adds even more concern.

A report by European-based Chinese researchers Fei Su and Jingdong Yuan, China's 'Thinking on Artificial Intelligence Integration and Interaction with Nuclear Command and Control, Force Structure and Decision-Making', reports that China has released two key position papers in recent years. They address AI and related risks: the 2021 Position Paper on the Regulation of Military Use of Artificial Intelligence and the 2022 Position Paper on Strengthening the Ethical Governance of Artificial Intelligence. The key findings of the researchers regarding these two documents are as follows:

1. Given the acceleration of China's integration of AI into military systems, it is unlikely that China will propose any regulation that would limit the development of military AI.
2. The documents outline opposition to the use of AI to undermine China's sovereignty and territorial security. This is a red line in China's position on the use of AI.
3. The importance of human intervention in weapons systems is emphasized – AI must always be under human control.

Taking into account the peculiarities of management methods at the state level, it can be argued that the danger of China's ambitions in the application of AI in the military sphere is quite real.

In our opinion, it is also important to pay special attention to the increasing interest in the use of AI in the military sphere in countries such as Iran and North Korea. Russia, which has been using AI elements in the war with Ukraine since 2022, announced the creation of an AI unit in the Department of Innovative Development of the Ministry of Defense in the same year. Trends in Russia's use of AI in the military sphere are rapidly accelerating.

 

Ukraine

Ukraine, at the level of individual projects, both cooperating with the world's leading companies and applying its own developments, is extremely actively using AI tools during combat operations against the Russian army.

As the Minister of Digital Transformation Mykhailo Fedorov notes, "Today we are a kind of training ground for the use of artificial intelligence. It is actively used for recognition and decoding of satellite images, recognition of targets on the battlefield, development of target capture technologies for drones, face recognition, and prediction of means by which certain targets can be hit."

Discussing the issue of AI regulation in the defense sphere, the Deputy Minister of Digital Transformation of Ukraine for IT Development Oleksandr Borniakov noted that during a full-scale war, AI regulation in projects related to military needs is not foreseen. As a delicate comment, it is worth paying attention to the fact that state regulation "for limitation" is only one of the tools of state management. In those countries where the principle of "good governance" applies, state regulation is used mainly "for encouragement" and long-term strategic and tactical planning.

Read this article in Ukrainian and russian.