UNIDIR’s Global Conference on AI, Security and Ethics 2025 provided a unique platform to address governance challenges and opportunities surrounding artificial intelligence (AI) in security and defence. The forum brought together nearly 500 participants, including academia, civil society, industry leaders, and research institutions, aiming to foster meaningful dialogue and collaboration.
The international community is grappling with the rapid advancement of AI technologies, including their implications on global, regional and national peace and security. While new AI systems and applications offer opportunities both in the military and security domains, their development, deployment and use bring legal and ethical considerations, as well as risks.
Policymakers and regulators worldwide have begun to recognize the urgency of shared understandings, norms and standards that transcend national borders and individual interests, especially in the context of AI’s applications for security and defence. This highlights the need for a platform to enable meaningful engagement with State and non-State stakeholders.
Shaping responsible AI governance
In late 2024, the UN General Assembly First Committee approved resolution on military AI governance. The resolution builds on the growing interest of States and the wider international community to assess the implications of AI in the military domain. It calls for submissions from States and stakeholders to address the opportunities and challenges AI poses to international peace and security.
In parallel, the international community has been deliberating on lethal autonomous weapons systems – a very specific subset of military AI, notably within the context of the Convention on Certain Conventional Weapons and its Group of Governmental Experts. Efforts to expand discussions beyond the convention have led to two resolutions adopted by the UN General Assembly in 2023 and 2024, emphasizing the need for responsible governance in this space.
The UN Secretary-General’s agenda has also prioritized the impact of emerging technologies in global security. Notable initiatives include the High-Level Advisory Board on AI; a resolution advocating for safe, secure and trustworthy AI for sustainable development, and a resolution promoting international cooperation in AI capacity-building. Commitments laid out in the Pact for the Future also include advancing discussions around lethal autonomous weapons and assessing the existing and potential risks associated with the military applications of AI.
How UNIDIR is driving AI security and ethics
To support these global efforts, UNIDIR launched the Global Conference on AI, Security, and Ethics 2025, a flagship event under the Roundtable for AI, Security, and Ethics, in partnership with Microsoft. This initiative aims to facilitate cross-regional and multisectoral dialogue, fostering cooperation across geopolitical divides.
The conference provided a platform for multistakeholder cooperation and ethical governance of AI, aiming to ensure its responsible development and deployment in security and defence. A message from the Secretary General, in support of the conference, was read during the opening by Izumi Nakamitsu, Under Secretary General and High Representative for Disarmament Affairs.
Over two days, the conference presented insights from over 50 global experts and included keynotes from leading voices in diplomacy, defense, technology, and humanitarian sectors. Key themes included AI’s role in military applications, cybersecurity, international law, human rights, and disarmament.
The programme facilitated interactive exchanges, lightning talks, and deep dives and featured high-level panels with UN and tech leaders on AI’s role in peace and security. Sessions explored AI-induced destabilization and security risks, as well as include presentations on AI in UN peacekeeping. A regional perspectives panel also addressed AI governance across different geopolitical contexts.
The Global Conference on AI, Security, and Ethics 2025 served as a cornerstone event in shaping the international governance of AI in security and defence. By bringing together diverse stakeholders, UNIDIR is committed to fostering open, independent and evidence-based discussions to navigate the evolving landscape of AI governance.
Latest insights from UNIDIR’s research
UNIDIR’s Security and Technology Programme is documenting where the points of convergence and disagreement are on many policy aspects of AI in security and defence. It provides expertise and support for States’ strategy and policy development, with key publications on AI including:
- Governance of Artificial Intelligence in the Military Domain: A Multi-Stakeholder Perspective on Priority Areas – Insights from government representatives, industry and the research community provided during the first convening of the Roundtable for AI, Security, and Ethics.
- AI in the Military Domain: A briefing note for States – As a follow up to the UN General Assembly resolution on AI in the military domain, this brief includes contextual information on the topic, a set of considerations for States to refer to, and a list of suggested readings that draws on UNIDIR’s own research and selected external publications.
- The Global Kaleidoscope of Military AI Governance – Key outcomes of the five regional consultations around responsible AI in the military domain held in 2024, in the lead-up to the second Responsible AI in the Military Domain Summit.
- Draft Guidelines for the Development of a National Strategy on AI in Security and Defence – An opportunity for States and all relevant stakeholders involved in the development, adoption, implementation and review of national strategies on AI in security and defence to contribute to policy development.
- Artificial Intelligence Beyond Weapons: Application and Impact of AI in the Military Domain – An exploratory report detailing 18 more applications of AI in defence to manage tasks other than deployment, as part of weapon systems.
- Gender and Lethal Autonomous Weapons Systems – Explores potential challenges to responsible implementation of AI in the military domain, such as gender bias.
