IEAI at Munich Security Conference 2026 Side Event

On February 13th, the IEAI had the pleasure of co-hosting with Amerikahaus an official side event to the Munich Security Conference 2026 entitled: “Ctrl-Alt-Deterrence: Rethinking Stability in the Age of Cyber, AI and Autonomy”. The panel brought together leaders from defense, academia, policy and industry to explore how artificial intelligence and autonomous systems are reshaping deterrence theory and practice.

The discussion was moderated by Brigadier General David Barnes, PhD (US Army, Retired) (Empowering AI) and Lance Lindauer (Partnership to Advance Responsible Technology). The panelists included Prof. Dr. Christoph Lütge (Director, TUM IEAI), Lauren Kahn (Senior Research Analyst, Georgetown’s Center for Security and Emerging Technology), Maryna Tymchenko (Director of Business Development and Operations, Blue Arrow Defense), and Felix Rank (Co-Founder, German Forum for Ethical Machine Decision Making).

The moderators opened the discussion by posing a central question: Does AI fundamentally transform deterrence, or does it instead accelerate existing strategic dynamics? As AI-enabled operations, drones and algorithmic decision-support systems compress timelines and complicate attribution, traditional assumptions about credible threats, escalation and stability are increasingly under strain. This evolving landscape raises important questions not only about technological capability, but also about responsibility, accountability, governance and the human role in decision-making.

Ms. Kahn suggested framing AI solely as either a revolutionary or incremental change risks oversimplifying its true impact. Rather than replacing deterrence theory outright, she emphasized that AI complicates deterrence by introducing opacity, automation bias and new escalation risks. Because AI systems are ultimately human-designed and deployed, she argued policymakers should focus on identifying specific vulnerabilities, such as accidents, misinterpretation of automated outputs and uneven technological adoption, and adapt governance mechanisms accordingly. In this sense, the challenge lies not in abandoning existing frameworks, but in refining them to address new operational realities.

Drawing on his military experience, Dr. Barnes underscored the importance of maintaining human judgment in increasingly automated environments. He highlighted how military planning has traditionally relied on careful deliberation, clear commander intent and defined chains of accountability. However, AI-enabled systems operate at machine speed, raising concerns about whether human operators will retain meaningful situational awareness and the ability to intervene when necessary. He emphasized that while AI can enhance operational effectiveness, it must remain aligned with mission objectives and human oversight, particularly in high-stakes environments where rapid decisions can have irreversible consequences.

From the perspective of operational deployment, Ms. Tymchenko provided insight into the realities of integrating AI-enabled systems in active conflict settings. She noted that while AI can assist with navigation and target identification, human operators remain responsible for final decisions regarding the use of force. In frontline environments, where personnel must act under intense pressure and uncertainty, there is strong resistance towards using fully autonomous lethal systems. Still, Ms. Tymchenko cautioned adversaries may not adhere to similar ethical constraints, creating asymmetries that complicate both strategic planning and defense.

The discussion then turned to the question of responsibility and accountability in human/machine teaming. Mr. Rank emphasized moral and legal responsibility must remain with humans, as code-developing systems cannot be held accountable. However, as AI becomes more deeply integrated into operational and organizational processes, traditional chains of responsibility and liability may become less clear. He stressed the importance of education and literacy across technical, military and policy to ensure decision-makers understand both the capabilities and limitations of AI systems.

Prof. Lütge expanded on the ethical dimensions of AI deployment, stressing ethics must be grounded not only in abstract principles but also in technical robustness and reliability. Different applications of AI require different forms of oversight, and meaningful human control depends on clearly defining when and how humans can intervene in AI-driven processes. Ensuring transparency, system accuracy and clear accountability structures is essential for building trust in these technologies, particularly in contexts where decisions may have life-or-death implications.

As the discussion broadened, panelists examined the structural implications of AI for global stability. While AI is often described as an “arms race”, several speakers cautioned against relying too heavily on analogies with nuclear deterrence. Unlike nuclear weapons, AI is a general-purpose technology with diffuse supply chains, widespread commercial applications and significant verification challenges. Rather than focusing exclusively on competition, the panelists emphasized the importance of transparency, shared standards, cooperation among allies and ongoing dialog among policymakers, industry and technical experts.

Questions of institutional adaptation and collective deterrence also featured prominently. Panelists noted traditional governance frameworks may struggle to keep pace with rapid technological change, while informal networks and partnerships may provide more agile mechanisms for coordination. At the same time, effective deterrence in the AI era requires interoperability, shared planning and a clear understanding of how AI systems interact with human decision-making structures. Dr. Barnes wrapped up the panel with a quote from former US President, Dwight D. Eisenhower: “Plans are worthless, but planning is everything”, encouraging a continued discourse.

The evening concluded with an engaging Q&A session, during which audience members raised questions about the balance between speed and control in autonomous systems, the role of AI in disinformation detection and the future of international governance. The IEAI thanks Amerikahaus for co-hosting this event, the panelists and audience for their thoughtful contributions, and the Munich Security Conference for supporting this timely and forward-looking discussion on deterrence and stability. The recording of the event can be found below.