Even if there is no Legal definition of AI, existing ones can provide guidance to policy makers
Hill and Marsan, 2018 - Director and Senior Assistant, NATO Office of Legal Affairs [Steven and Nadia, 7-18-18 “Artificial Intelligence and Accountability: A Multinational Legal Perspective” https://www.sto.nato.int/publications/STO%20Meeting%20Proceedings/STO-MP-IST-160/MP-IST-160-PP-4.pdf Acc 4/21/22 TA]
2.0 DEFINITIONS The first issue that arises in the context of AI which highlights the difficulty in setting a clear conceptual scope and applicable legal parameters, is a definitional one. There are no internationally agreed legal definitions for the core concept of AI. The lack of agreed legal definitions can hinder and stall discussions in part because Nations may rightfully be hesitant to commit to the regulation of a new technology when the scope and the evolution of that technology is not clear. While there might not be agreed legal definitions of AI, broader discussions on AI have identified a number of elements that can nevertheless provide some preliminary guidance. The first suggested definition of AI is, “the capability of a computer system to perform tasks that normally require human intelligence, such as visual perception, speech recognition and decision-making.”6 The second definition of AI which is useful here is, “technologies that enable machine learning, natural language processing, deduction through vast data-computational power, and ultimately, automated decision-making in robotics or software that can substitute for tasks once performed exclusively by human action and judgement”. 7 Another term that is important to define is “autonomy”. The concept of autonomy is key to AI because it is precisely the technological edge provided by AI that enables autonomy or, expressed another way, enables the independent action of a machine. Autonomy itself can be defined as “the ability of a system, platform or software to complete a task without human intervention, using behaviours resulting from the interaction of computer programming with the external environment”.8 Although beyond the scope of this paper, other definitions incorporate key concepts in the military use and development of AI.9 These preliminary definitions suggest that AI is about replicating human perception, cognition and decision-making as well as introducing a certain element of independence to these systems. Although the modelling of human intelligence provides opportunities from a security perspective, these are also fraught with challenges that are only beginning to be understood: “there is now a broad consensus that AI research is progressing steadily, and that its impact on society is likely to increase. The potential benefits are huge, since everything that civilization has to offer is a product of human intelligence; we cannot predict what we might achieve when this intelligence is magnified by the tools AI may provide, but the eradication of disease and poverty are not unfathomable. Because of the great potential of AI, it is important to research how to reap its benefits while avoiding potential pitfalls”. 10 3.0 LEGAL IMPLICATIONS ARISING FROM SPECIFIC USES OF AI There are already applications of AI that are of interest to a multilateral security organisation such as NATO. Acknowledging the absence of a clear consensus on the legal definition of AI, it is useful to highlight some concrete areas where AI has been and is currently being used with effects, both positive and negative, on Allied security.11 Three areas of AI-enabled technologies are worth mentioning here: intelligence, surveillance and reconnaissance; the manipulation of personal data; and, disinformation.