By Nida Shahid
For a long time the term Artificial Intelligence (AI) evoked images of ‘Terminator-style’ killer robots and machines eventually taking over the human world similar to ‘The Matrix’. It is only in the last decade that governments, private industries and academia have begun pouring substantial investments and research in this domain to better understand its nuances. However, AI’s potential role in key military strategies, such as deterrence, remains a relatively underexplored area.
According to experts, the integration of AI and deterrence is likely to bring about another revolution within the national security landscape. Imagine a world where decision-makers have access to real-time data with the efficiency of AI to analyse this information within seconds, leading to identification and neutralisation of threats with unprecedented speed and accuracy. This could also result into a security landscape where the risks of accidental escalation are significantly reduced. At least, that is the promise of AI’s amalgamation with the concept of deterrence.
Addressing these complexities requires decision-makers to be more mindful to ensure responsible leveraging of AI technologies, especially for deterrence purposes, while aligning with established international and humanitarian norms. This approach will be crucial in maximising the benefits of AI for augmenting deterrence while mitigating potential risks
On the flip side, however, there are profound challenges, including ethical dilemmas which accompany AI’s ascent in this domain. As states move forward in their technological pursuits, decision makers will have to contend with not only the advantages of leveraging this new technology, but its potential consequences as well.
One of the core concepts of deterrence is that stronger defensive capabilities will lead to more effective ‘deterrence by denial’ options, which is considered to be the preferred deterrence model. In this regard, AI has the potential to bolster precision, speed and reliability of defensive weapon systems. By doing so, it can dissuade potential aggressors from pre-emptive options.
Beyond kinetic actions, AI can also help improve states’ abilities to enhance the planning and execution of strategies for both offensive and defensive purposes. AI can help states enhance their logistics, coordination, training, communication and deployment, thereby significantly improving the overall effectiveness of military operations. As highlighted by the on-going conflict in Ukraine, AI-driven back-office systems governing the complexities of warfare can make coercive threats more robust, persuasive, and feasible.
Moreover, AI can enhance the overall credibility of deterrence. By providing accurate and timely intelligence, AI systems can support more informed decision-making, thereby strengthening the deterrent posture. This capability is particularly important in the context of nuclear deterrence, where the stakes are exceedingly high.
Despite these clear advantages, the integration of AI into deterrence strategies is not without its challenges. Innovations in warfare technologies and methodologies often lead to adversaries mimicking actions and AI-enhanced decision-making is likely to be no exception. If one side in a conflict gains significant advantages owing to integration of AI, the other side is also likely to adopt similar technologies. This could lead to a future where competing AI systems, operating at machine speed offer only brief windows of advantage, leading to a use-it-or-lose-it mentality among adversaries and altering traditional escalation calculations.
The opacity of AI algorithms presents another significant challenge. Currently, there is a lack of transparency pertaining to how AI systems actually make decisions, which can lead to mistrust among states. Even among adversaries with deliberately ambiguous military or nuclear policies, there is an understanding of each other’s thresholds and redlines. Without a clear understanding of an adversary’s decision-making processes, particularly those involving AI systems, the risk of misinterpretation and unintended escalation increases. Dr James Johnson, a specialist in AI and its strategic applications, highlights that AI’s propensity for unpredictability could undermine strategic stability rather than enhance it, especially if adversaries perceive decisions taken by AI as irrational or overly aggressive.
In addition, AI systems, while highly advanced, are not infallible. They are susceptible to biases, errors, and cyber-attacks, which could lead to unintended escalation. Moreover, the use of AI in deterrence postures also raises ethical and legal questions. Delegating critical decision-making to AI challenges not only the ethical norms of deterrence but also the legal frameworks and norms governing the use of force. The potential for AI to autonomously initiate military actions blurs the lines of accountability and responsibility, raising concerns about compliance with international humanitarian law.
Addressing these challenges requires multifaceted approaches. Establishing robust AI governance and accountability frameworks at the international level is crucial. This includes promoting transparency in the use of AI algorithms for deterrence purposes and ensuring that human oversight remains a cornerstone of critical decision-making in the military sphere.
Addressing these complexities requires decision-makers to be more mindful to ensure responsible leveraging of AI technologies, especially for deterrence purposes, while aligning with established international and humanitarian norms. This approach will be crucial in maximising the benefits of AI for augmenting deterrence while mitigating potential risks.
The writer is a Senior Researcher at Centre for Aerospace and Security Studies Lahore. She can be reached at [email protected]
