Photo. Fot. Rufio1/Pixabay
Copy link
Send email
This piece is based on GLOBSEC’s latest brief titled Leveraging Artificial Intelligence for NATO’s cyber resilience: Preliminary perspectives, published as part of the GLOBSEC Future of Cyberspace Initiative: Transatlantic Chapter. Authors: Michael Chertoff and Anushka Kaushik
Russia and China, among others, are using Artificial Intelligence (AI) to improve their efforts to penetrate U.S. and allied networks, conduct network-based espionage, and wage malware campaigns. NATO must raise its defences by following suit. Developments in AI-augmented security are promising. New tools are enabling defenders to optimise processes, rapidly evolve defenses, analyse large data sets, and automate repetitive cybersecurity tasks. Moreover, Alliance members have already made progress in applying AI to threat detection, response automation, and resilience-building.
However, to enable NATO to best leverage AI to bolster its cybersecurity and counter these threats, the Alliance needs to work on a broader strategy with a clear outline of challenges and opportunities – such a strategy would include addressing concerns around interoperability, data sharing and access, and inculcating a cautious approach to the application of technology in defence environments.
A primary concern is interoperability – at the core of the interoperability challenge is the notion that Allies might use different or siloed AI systems that cannot collaborate to counter cross border or multi-vector cyber threats. Interoperability is critical to the Alliance’s cybersecurity efforts as it enables architecting AI solutions that can work across multiple allies and models. While NATO is making progress, it remains difficult to achieve.
Secure data sharing, real-time analysis of large datasets, and the timely processing of information are all necessary to make interoperability a reality. Equally crucial is the standardisation of data access and usage, as even minor variations can distort analytical outcomes. This is compounded by differences in national data governance policies, making it difficult to assemble coherent datasets for collective analysis. Synchronising these rules across Allies for the purpose of consistency and to avoid analytical distortions will be crucial.
Data also presents tremendous opportunities for facilitating a robust understanding of where cyber threats come from. One such example is enhancing anomaly detection by using a multidomain approach that allows analysis beyond a single layer of a system. This means looking at data from multiple domains across the supply chain ecosystem which can have advantages in anticipating where cyber-attacks are emerging from. Additionally, monitoring software and hardware supply chains to detect anomalies within these datasets that might not be apparent through individual network analysis is an area of opportunity.
NATO should continue to make progress in bridging the knowledge gap. Specifically, strengthening engagement with the private sector and enhancing cross-sector collaboration with governments and research institutions are essential steps—initiatives such as NATO’s Science for Peace and Security Programme provide a useful platform for this purpose.
NATO’s AI strategy should also integrate stronger political messaging, ensuring decision-makers understand the strategic importance, ethical use, and deterrence effects of AI technologies in military contexts. This should also cover the ethical implications of using autonomous systems vis-à-vis the operational benefits so that informed policy decisions can be taken.
Underlying all these should be a cautionary approach to AI – a clear assessment of what the risks are, how the Alliance can assess these risks, and the how Allies will determine the maturity of existing AI tools. Caution in deployment, assessment of maturity levels, and reliability standards in defence environments will need to be prioritised. Moreover, a case-specific approach for AI tools that are still in the nascent stages of evaluation should be undertaken.
NATO is navigating an evolving cyber threat landscape amidst rising geopolitical instabilities. In order to stay ahead, it must leverage the development in AI to bolster its cyber resilience. AI-driven capabilities can strengthen anomaly detection, automate threat response, and provide deeper insights into emerging cyber threats. Even as NATO has to navigate the various challenges that must be faced in AI adoption, it must also prioritise creating a foundational baseline of capabilities across all member states, and cultivating strategic thinking around the technology.