1. NATO standards are ineffective in establishing interoperability on AI.
Erica Pepe, 2020 (Senior Coordinator for Research and Conflict, Security and Development Analyst, “NATO and collective thinking on AI,” https://www.iiss.org/blogs/military-balance/2020/11/nato-artificial-intelligence, Retrieved 6/15/2022)
NATO could play an important role in establishing interoperability standards and norms of use in the military application of artificial intelligence. But will allies be able to adopt a common approach in the use of AI for defence and security and keep the Western technological edge? By the end of 2020, NATO may have in place broad thinking on artificial intelligence (AI) as part of a wider implementation strategy on how to benefit from emerging technologies. The Alliance could play an important role in establishing interoperability standards along with supporting norms of use in the military application of AI. NATO, however, has not always been successful in establishing standards, nor in getting member states to comply. This makes the recently set-up NATO Innovation Board, and its intended year-end emerging-technologies implementation strategy, potentially all the more important. An AI White Paper, including a proposed five-year roadmap, is already with Alliance governments, having been submitted in July. This could lead to a more detailed NATO AI strategy emerging sometime in 2021. The AI challenge The Alliance has the advantage of having 30 members spending some US$1 trillion (2020) on defence. Its wide membership, however, is also a possible weakness if even a top-level common approach to AI cannot be agreed. The Innovation Board’s chair, NATO Deputy Secretary General Mircea Geoană, has the task of coordinating ongoing AI-related work across the Alliance in attempting to shape a shared vision of how to utilise AI. This will include working to identify capability needs and potential projects for collaboration. NATO already has some AI-related projects under way, including the Military Uses of Artificial Intelligence, Automation, and Robotics project and the NATO Data Science Centre. The former project is part of the United States-led Multinational Capability Development Campaign and is intended to look at possible applications in areas such as air and missile defence. The Data Science Centre’s aim is to bring together existing data-science expertise within NATO under one structure. The Alliance views AI as one of several of what it calls ‘emerging and disruptive technologies’, which it considers will likely have the greatest impact in combination. NATO’s March 2020 ‘Science & Technology Trends 2020-2040’ report pointed to the mix of autonomy, big data and AI, for example being used to generate decision-relevant information from vast quantities of raw data, as having far-reaching implications for defence and security. The report also identified an array of ‘serious challenges’ that NATO faces in managing AI, including policy, legal, ethical and interoperability issues. It further identified the need for a common approach to how the performance of AI-software-based systems is verified, and in turn how the verification process is itself validated. In some ways NATO might seem a natural forum for these deliberations, not least in a transatlantic context. It also has a lot of experience, going back to the Cold War, in working towards standardisation and interoperability among allies. However, the results achieved have been mixed, which underscores the challenges the Alliance now faces: not only 30 members with disparate levels of capability, but also a backdrop of rapid technological advances where some of its competitors and potential adversaries may hold significant advantages. Interoperability issues may be thorny, but they need to be resolved if AI-dependent capability gaps between members are not to widen. Early discussions regarding the establishment of common technical standards on the design and development of military-applicable AI would at least reduce this risk.