AI in Military Systems: Conversations from the Munich Security Conference
Rohan Barad | 14 March 2025
Summary
The Munich Security Conference highlighted a focus on the AI arms race amongst major powers, with advancements in autonomous weapons systems.
Implications include potential disruption of strategic stability and a heightened risk of conflicts.
Forecasts suggest a rapid spread of AI military technologies, increased international tensions, and possible arms control talks.
The Munich Security Conference 2025 brought together world leaders, defence experts, and policymakers to discuss pressing global security challenges. A key focus was the rapid advancement of AI-powered military systems by major powers, which has become a significant and intricate issue requiring immediate consideration.
The integration of AI into military systems has critical consequences for global security. As AI systems become more autonomous, there's a growing risk of miscalculations or system errors leading to unintended military engagements. The UK Technology Secretary, Peter Kyle, highlighted that state-sponsored hackers are already using AI to write malicious code and identify system vulnerabilities, increasing the sophistication and efficiency of their attacks. Moreover, criminals are employing AI deepfakes to assist in fraud, breaching security by impersonating officials. In one instance, attackers used live deepfake technology during a video call to mimic bank officials, resulting in a theft of USD 25 million.
The conference also highlighted a shift in focus from last year's concerns about trust, safety, and potential misuse of AI in election manipulation to the opportunities AI offers. Discussions at events like COP30 are also highlighting the need for sustainable AI practices across all sectors, as the energy consumption of AI systems becomes a growing concern. This dual nature of AI as both a tool for climate action and a potential contributor to emissions adds another layer of complexity to the issue.
The widespread preference for bilateral deals rather than inclusive multilateral cooperation, as revealed in the Munich Security Index 2025, underscores the challenges ahead in fostering global collaboration on AI governance and security. a trend exemplified by several concrete outcomes. For instance, the UK's Sovereign AI Strategy Britain unveiled its bilateral agreement with Anthropic to co-develop AI systems for "national security priorities" while rebranding its AI Safety Institute as the AI Security Institute focused on "protecting UK interests".
The U.S. National Archives
Forecast
Short-term
Calls for international discussions on AI arms control and proposals for multilateral talks will likely increase to address the risks associated with AI in military applications.
Long-term
Investment in AI defence technologies will likely increase, including AI integration in the command system and training. This could lead to the popularisation of AI-enhanced military capabilities beyond the major powers.