14. Artificial intelligence and international peace and security

Advances in artificial intelligence (AI) are poised to bring enormous benefits, but they can also create or exacerbate existing threats to international peace and security. In recent years, states have increasingly acknowledged the need to manage these complex risks—stemming from both civilian and military AI—through the establishment of new forums and initiatives. Discussions and work related to these forums and initiatives continued in 2025.

Military AI

For the past decade the international policy conversation on military uses of AI has mostly focused on autonomous weapon systems (AWS), commonly characterized as weapon systems that, once activated, can select and engage targets without human intervention. Since 2023, however, the conversation has expanded to other military uses of AI such as in targeting, planning and intelligence analysis. Reported uses of AI in major armed conflicts in 2025, especially Gaza and Ukraine, led policymakers to pay particular attention to so-called AI-enabled decision support systems.

Discussion on AWS has historically primarily taken place in Geneva under the framework of the 1981 Convention on Certain Conventional Weapons (CCW Convention), but in 2025 the United Nations General Assembly started a dedicated discussion track in New York. In addition, the UN secretary-general, António Guterres, published his first-ever report on AI in the military domain, and the UN General Assembly agreed to convene informal exchanges in 2026 on the peace and security implications of military AI. The three-year mandate of the CCW Convention’s group of governmental experts concludes in 2026 and next steps regarding the regulation of AWS will need to be agreed—not least whether states can start formally negotiating the framework and content of a new governance instrument.

Civilian AI

Efforts to understand the implications of civilian AI for peace and security also continued to evolve during the year. Developments included an agreement by UN member states to establish an Independent International Scientific Panel on AI; the creation of a multistakeholder Global Dialogue on AI Governance; and the AI Action Summit in Paris. In addition, the European Union and several states leading the development of AI—including China and the United States—transitioned from foundational laws or voluntary AI principles (mostly drafted in 2023–24) to adopting or enacting formal, enforceable governance processes and new detailed compliance guidelines and directives.