UN Launches Initiative to Address Military Uses of Artificial Intelligence

The United Nations is launching a new initiative to address the military applications of artificial intelligence, prompted by significant concerns regarding the technology's potential threat to humanity. The UN Institute for Disarmament Research (UNIDIR) plans to establish the first international forum dedicated to this issue, aiming to bring key players together to tackle the challenges posed by rapidly advancing AI capabilities.

Robin Geiss, head of UNIDIR, highlighted the shrinking margin for human judgment, even in critical areas such as nuclear operations, and stated that the uncertainty surrounding autonomous systems is a primary concern the initiative seeks to resolve. Geiss noted that AI is increasingly integrated into various systems, from early warning networks to intelligence gathering, making it difficult to determine when human control might be lost.

There has been a growing recognition that action is needed on military AI. This sentiment has been fueled by the proliferation of AI capabilities and their potential global impact. Concerns about AI's influence are reportedly transcending geopolitical divides, prompting actions from major world powers. For instance, a recent proposal suggested the United States and China establish an AI safety alert system for issues that could jeopardize national security.

The risks associated with AI were a prominent topic during the UN's annual high-level meeting last month, with leaders from many of its 193 member nations addressing the issue in their speeches.

UNIDIR established a Center of Excellence on AI, Peace and Security earlier this year, serving as the world body's first permanent platform focused on these subjects. The center has received funding from the five major nuclear powers: the United States, Russia, China, Britain, and France.

As part of this center, UNIDIR is launching a new forum to convene diverse stakeholders from across sectors and the globe to collaborate on AI's military uses. The initiative will also focus on the role of the industry, aiming to develop a framework for responsible industrial behavior. Geiss emphasized that corporate decisions in AI design and development are crucial for ensuring the reliability, predictability, and adherence to international law of AI-enabled military capabilities.

UNIDIR plans to conduct consultations with defense technology companies, software developers, investors, and other industry representatives. These consultations will involve governments, civil society, academia, and experts. Furthermore, UNIDIR is initiating a project to address the risks of AI misuse by armed groups, noting that such entities are already employing AI for operational planning, execution, and logistics.

Geiss concluded by stating that the urgent issue of AI in the military domain requires a consolidated effort to confront the challenges directly.