NextFin News - In a significant blow to international efforts to regulate the next generation of warfare, the United States and China both abstained from signing a joint declaration on the responsible use of artificial intelligence (AI) in the military domain. The summit, held in A Coruña, Spain, on February 5, 2026, concluded with only 35 of the 85 participating nations endorsing a set of 20 voluntary principles. While major allies including Canada, Germany, France, the United Kingdom, and South Korea signed the document, the absence of the world’s two primary AI superpowers underscores a growing global divide over the ethics and control of autonomous weaponry.
The declaration, drafted during the third Responsible AI in the Military Domain (REAIM) summit, sought to establish a framework for human accountability, clear chains of command, and rigorous testing for AI-enabled systems. According to Reuters, the document emphasized that humans must remain responsible for the deployment of AI weapons and encouraged the sharing of national oversight mechanisms. However, unlike previous summits in The Hague (2023) and Seoul (2024), where the U.S. had previously supported a looser "blueprint for action," the 2026 Spanish summit saw Washington join Beijing in the ranks of the non-signatories, reflecting a hardening of strategic positions under the administration of U.S. President Trump.
The decision by U.S. President Trump’s administration to opt out appears rooted in a desire to maintain "strategic ambiguity" and avoid any constraints that could hinder rapid technological iteration. Analysts suggest that Washington is increasingly wary of binding or even semi-formal international rules that might erode its first-mover advantage. By avoiding these standards, the U.S. preserves its ability to develop and deploy sensitive battlefield AI without the transparency requirements that the declaration encouraged. This shift also reflects a broader skepticism toward multilateral frameworks that include non-allied nations, with the U.S. preferring to set standards within its own "small-circle" of trusted partners.
China’s refusal to sign, articulated by Li Chijiang, Deputy Director-General of the Department of Arms Control, followed a different rhetorical path but led to the same outcome. Li emphasized a "human-centered" approach and cautioned against an "obsession with absolute military advantage," yet Beijing remains deeply suspicious of Western-led regulatory frameworks. According to a readout from the Chinese Ministry of Foreign Affairs, China views many of these international declarations as potential tools for technological hegemony that could constrain the development of non-Western states. Furthermore, as China rapidly closes the gap in military intelligentization, it is unwilling to accept oversight mechanisms that it perceives as ideologically biased or strategically restrictive.
The resulting landscape is what Dutch Defence Minister Ruben Brekelmans described as a classic "prisoner’s dilemma." In this framework, nations are caught between the rational desire to prevent a catastrophic, unintended AI escalation and the fear that self-imposed restrictions will leave them vulnerable to adversaries who do not follow suit. Brekelmans noted that Russia and China are moving at a pace that creates an "acute necessity" for Western development, effectively making the arms race the primary driver of policy, rather than the humanitarian guardrails discussed in Spain.
From a financial and industrial perspective, this lack of consensus is likely to accelerate capital expenditure in the defense-tech sector. With no universal "rules of the road," the U.S. and China are expected to double down on AI-integrated hardware, from autonomous drone swarms to algorithmic decision-support systems. This trend is already visible in the market; major U.S. defense contractors and specialized AI firms are pivoting toward "black box" development cycles that prioritize speed and lethality over the transparency requested by the REAIM signatories. The failure of the Spain summit suggests that the future of military AI will not be governed by a single global treaty, but rather by competing technological blocs, each operating under its own internal ethics and operational standards.
Looking forward, the trend toward fragmented governance appears irreversible in the near term. As AI becomes more deeply embedded in national security infrastructure, the difficulty of verification and enforcement makes any non-binding declaration increasingly symbolic. The 35 nations that did sign—largely middle powers and European states—may find themselves in a difficult position, attempting to uphold high ethical standards while their primary security guarantors or rivals operate without such constraints. The A Coruña summit may well be remembered as the moment when the window for a unified global approach to AI warfare officially closed, replaced by a high-stakes race for algorithmic supremacy.
Explore more exclusive insights at nextfin.ai.
