The greatest emerging threat to international peace and security is the integration of artificial intelligence (AI) into nuclear command, control, and communications (C3) systems. In particular, this applies to the use of AI in early-warning detection systems at the precise geopolitical moment when binding arms control agreements are eroding. For decades, nuclear stability rested on deterrence reinforced by formal treaties. Today, those guardrails are gone: the Intermediate-Range Nuclear Forces Treaty collapsed in 2019, the Open Skies Treaty soon followed, and the New START bilateral agreement expired in February 2026.
Alongside this regulatory decay, a volatile nuclear landscape has emerged. The traditional U.S.-Russia duopoly has given way to a multi-polar environment featuring nations like India, China, North Korea, and Pakistan, while Iran remains a threshold state of significant concern. This expansion of nuclear stockpiles compounds institutional fragility, unfolding alongside rapid advances in AI that threaten to destabilize strategic decision-making and crisis stability.
The Encroachment of AI into Defense Architectures
AI companies are becoming increasingly embedded within military frameworks. In the United States, firms such as OpenAI, Anthropic, and Google are deeply integrating into the Department of Defense. Similarly, in India, Bharat Electronics Limited has assumed a central role through Akashteer, an AI-powered command-and-control platform.
These systems leverage AI to fuse satellite imagery, radar returns, and multi-source sensor data to support decision-making and targeting. In the nuclear domain, this trend intersects with C3 systems to detect and classify potential missile launches. However, the use of AI is expanding beyond decision-support functions, increasing the risk that machine outputs could significantly shape escalation decisions.
[Traditional System] —-> Human Sensor Analysis —-> Deliberate Crisis Response
[AI-Fused C3 System] —-> Automated Data Fusion —-> Accelerated Escalation Pressure
Recent wargaming highlights this vulnerability. A study by Professor Kenneth Payne of King’s College London examined how modern AI models (including GPT-5.2, Claude Sonnet 4, and Gemini 3 Flash) simulated nuclear crises. Across 21 scenarios, the models displayed sophisticated escalation logic consistent with classical strategic theory, yet nuclear signaling occurred in almost all games and no model ever chose outright concession. Nuclear escalation was near-universal: 95% of the games saw tactical nuclear use and 76% reached strategic nuclear threats. Models like Claude and Gemini treated nuclear weapons as legitimate instrumental tools rather than moral thresholds. Under severe time pressure, AI systems consistently produce nuclear escalation-prone outcomes.
A Blueprint for U.S.-India Coordinated Governance
Without institutional guardrails, the reduction of human judgment in response scenarios risks accelerating miscalculation. Addressing this challenge requires coordinated leadership, specifically between the United States and India. As Former Indian Ambassador Bisaria indicated at the 2026 NXT Fellowship, India is uniquely positioned to lead multilateral coalitions capable of bridging the gap between major nuclear powers and emerging AI states.
To ensure AI integration yields a net benefit rather than a nuclear catastrophe, the United States and India must collaborate on a multi-pronged governance framework:
-
Establish an International AI-Nuclear Governance Body: Championing a global body tasked with setting international standards and updating norms as AI capabilities evolve.
-
Codify Worldwide AI-Nuclear Human Control Standards: Initiating a multilateral agreement ensuring AI is restricted to supportive detection analysis, mandating that humans retain absolute, final decision-making authority over any nuclear launch.
-
Negotiate Global AI-Nuclear Arms Treaties: Jointly advocating for a modern successor to the New START treaty that focuses squarely on AI-enabled weapons systems and automated early-warning infrastructure.
The greatest emerging threat to international peace is neither AI nor nuclear weapons in isolation. It is their convergence at a moment when institutional safeguards are weakening. Overcoming this existential vulnerability requires the United States and India to act decisively, collaborating with the international community to ensure technological innovation does not outpace human survival.
The author is from: Nina Gohel NXT Fellow 2026
