NATO AI governance solves hybrid tactics, run-away offensive cyber weapons, great power conflict, and internet functionality---diffusion of tech is inevitable but cooperation through NATO is key to standardization that leads to effective management
Sonia Lucarelli et al 21. Sonia Lucarelli is Professor of International Relations and European Security at the University of Bologna, and member of the Board of Directors of the Istituto Affari Internazionali (IAI). Alessandro Marrone is Head of the Defence Programme of IAI and teaches at the Istituto Superiore di Stato Maggiore Interforze (ISSMI) of the Italian Ministry of Defence. Francesco N. Moro is Associate Professor of Political Science at the University of Bologna and Adjunct Professor of International Relations at the Johns Hopkins University Europe Campus. “NATO Decision-Making in the Age of Big Data and Artificial Intelligence” Editors: Sonia Lucarelli; Alessandro Marrone; and Francesco Niccolò Moro. This publication is the result of the Conference “NATO Decision-making: promises and perils of the Big Data age”, organized by NATO Allied Command Transformation (ACT), the University of Bologna and Istituto Affari Internazionali (IAI) of Rome. https://www.iai.it/sites/default/files/978195445000.pdf //pipk
Digital revolution has substantially transformed the world we live in, providing great opportunities but also making societies more vulnerable and transforming the meaning of state borders. Technology makes external interferences cheaper, faster and all-encompassing: citizens can potentially become direct targets of information warfare, all members of a society can be part of conflicts one way or another. From advanced weaponry to command and control, most security-related domains are undergoing deep transformations as data availability and transmission increase exponentially. This is especially true as the emergence of so-called hybrid tactics contributes to universalize the battlefield. Also, attackers may lose control of their offensive cyber weapons, and ‘collateral damages’ across the private sector and the public worldwide might be more and more difficult to contain. Less visible, yet important challenges connected with Information Communication Technologies (ICTs) also exist. For instance, data overload can create problems for decisionmakers that are unable to detect important signals. Losing sight of how machines make their calculations – a somewhat inherent feature of Artificial Intelligence (AI) – can hinder deeper understanding of phenomena as well as learning, besides having dense ethical implications.
A crucial question for Western societies and governments is how to deal with technological changes by exploiting their many benefits while managing to limit their risks. Broadly speaking, observers have long noticed the potentialities of technologies in the security domain: better situational awareness, early warning against threats and risks, the ability to prevent and/or stop attacks to happen, the use of technology against the adversaries’ own technologies, and eventually deterrence of high-end hybrid warfare or, at least, the increase of resilience against it. In particular, in order to harness the potential of new technologies, higher levels of security are needed. While internet is unfortunately not secure by design, it has to be somehow retrofit to guarantee a certain level of protection – for instance by avoiding a single point of failure, developing better firewalls, etc. Ultimately, the digital revolution poses challenges to decision makers both as potential users of new technologies and as leaders of targeted societies. Learning to achieve political aims through the support of technological innovations and at the same time acquiring the ability to prevent and manage interferences, if not attacks, have become paramount.
However, achieving such results is not only about engineering. Technologies need ad hoc governance, organizations and skilled users to properly function. Actually, history is full of examples of good technologies that were improperly used and/or unable to provide the expected gains. Therefore, a joint, multi-disciplinary efforts is needed to think and manage technologies in a more comprehensive and secure way across various domains. For instance, the very same design of AI needs exchanges with social scientists in order to limit analytical biases and increase the quality of data that will then be processed through Machine Learning (ML). Moreover, many public policies involve technologies with a strong security dimension. This is one of the main reasons security standards should be harmonized across individual government’s policies as well as among Allies: this is what has been leading NATO’s renewed efforts on standardization beyond the strictly military perimeter, for instance towards the 5G domain.
While digital technologies continue to dramatically increase in scope and relevance, they are deeply embedded into the broader geopolitical framework, with the re-emergence of multipolarism and looming great power confrontation. This connection has to be discussed and understood as it affects not only security but also economic and technological domains. The globalized supply chain of technology building block entails vulnerabilities and dependencies on unreliable suppliers. Foreign Direct Investments (FDIs) in hightech companies, Small and Medium Enterprises (SMEs) and critical infrastructures are guided not only by an economic rationale but also by a politico-military one, and have to be monitored accordingly. Cyber space and, partly, outer space are de facto unregulated global commons where the ability to set regulations and standards could be a matter of competition and/or cooperation among major countries worldwide. The notions of ‘whole-of-government’ and ‘whole-of-society’ approaches confirm that these problems should be dealt with comprehensive strategies.
Great and middle powers increasingly rely on stand-off weapons, both physical and cyber ones, able to create damages rapidly, worldwide and on a large scale. This trend is going to be accelerated by AI. Some countries are adopting principles on responsible use of AI, including in terms of control and accountability. However, a vacuum remains in international law. And such vacuum is more difficult to fill because of the aforementioned interaction between geopolitics and technologies. Different powers conceive technology – and what it can bring them in terms of benefits – in different ways, and they are unwilling to regulate internationally this field of competition and warfare.
In such a rapidly changing security environment, NATO and allied activities directly or indirectly defend citizens’ daily life. In the age of Big Data, AI and the pervasive use of internet, the challenge is to defend the ever-expanding information environment while maintaining all its functionalities.
Against this backdrop, in the post-Cold War period NATO somehow missed the opportunity to involve Allies and partners in a debate on how defense technologies and norms have been changing with the ICT revolution. The result is that the web is not secure by design, and both private and public actors struggle to mitigate risks and threats in an unregulated environment where attackers are structurally advantaged over defenders. Today, the Alliance should not miss the opportunity twice vis-à-vis Big Data, AI and, broadly speaking, the current and future (r)evolution of ICT. The aim of the 2020 Academic Conference was precisely to explore some fundamental aspects of the challenges and opportunities posed by technological change to the security environment in which NATO works. Below follows a brief introduction to NATO, cyber defense and three sets of issues investigated in closer detail: Big Data and decision-making; hybrid threats to allied decision-making; AI adoption by allied armed forces.
NATO, Cyber Defense and Emerging Disruptive Technologies NATO begun to focus on cyber defense already in 2008, and over time it built up institutions and frameworks to deal with it from a well-limited military perspective. Allies recognized a cyber attack could lead to the activation of Article 5 of the Washington Treaty on collective defense. In that case, there is a clear procedure where NATO authorities take the military lead. Article 5 does not prescribe a clear procedure factoring in new technologies. On a regular basis, headquarters and the Secretary General cabinet carry on exercises on situational awareness, whereby they receive intelligence and military advice and are immersed in an information space with blue and red teams. Moreover, every two years, there is a large-scale exercise involving national governments. These efforts aim to build familiarity with the technology-related security challenges. However, further evolution of AI-based cyber attacks can constitute an increasing threat for datareliant organizations such as NATO.
Beyond cyber defense, the Alliance started to work on the broader issue of Emerging Disruptive Technologies (EDTs) only in 2019, by setting up an innovation board co-chaired by the Deputy Secretary General and the Supreme Allied Commander Transformation. Moreover, a dedicated unit was created in the Emerging Security Challenge Division. Two White Papers were produced, respectively on AI and on autonomous weapons, to provide inputs for Allies’ decisions in this regard. The current NATO approach is based on the motto “adopting and adapting”, entailing five complimentary goals: (1) better understand emerging disruptive technologies; (2) properly look at their implications for defense; (3) decide about their use; (4) mitigate their risks; and (5) exploit their advantages.
Noticeably, the traditional defense industrial ecosystem entails long planning, oligopolistic supply and monopsonic demand. Over time, it was characterized by substantial technology transfers from the military to the civilian domain (the so-called ‘spin offs’), including the very same embryonic Internet. In recent years, several new technologies with relevant implications for security and defense have been emerging from a different ecosystem, marked by bottom-up innovation, a rapid development-to-market cycle, and a technology transfer from the civilian to the military domain. As a result, with the relevant exception of certain space assets and hypersonic technologies, the civilian sector is increasingly developing into the innovation driver, and defense one has become quite dependent. Such a shift implies that priority setting for current and future technology development is not substantially driven by states anymore. In the US, the Pentagon’s Defense Advanced Research Projects Agency (DARPA) struggles to develop a dialogue with the private sector gravitating around the Silicon Valley to embrace certain research lines. The NATO Industry Partnership on the cyber domain serves as platform for Alliance’s officials and industrial representatives to exchange notes, yet major ICT players do not seem very interested in having such a structured dialogue. Moreover, investments in these technologies require venture capitals and the acceptance of the risks to fail – something which usually states, and particularly Ministries of Defense, cannot afford. The US, the UK, France, Germany, the Netherlands and other Allies made certain steps to adapt their defense innovation models in these domains, but this is only the beginning of a long transformation process.
As a matter of fact, adapting to emerging and disruptive technologies is harder for some Allies than others. The related risk is moving towards a multi-layer Alliance, with some member states holding new technologies, and others not having such advantage. Ideally, the solution would be to collectively adopt certain new technologies, but this represents a challenge for the NATO Defence Planning Process, military procurement, common funding, etc. A technology group of experts has been appointed to reflect upon issues including but not limited to these, and the Secretary General will probably present a report to the next summit of Heads of state and governments.