Rome Declaration Calls for Human‑Controlled AI in Nuclear Weapons

From July 14‑16 2026, the Vatican hosted the Global Nobel Laureates Assembly on Artificial Intelligence and Nuclear War at the Borgo Laudato Si’ gardens in Castel Gandolfo and the Capitoline Hill in Rome. Over 200 participants—including Nobel laureates, scientists, religious leaders and former heads of state—signed the “Rome Declaration for an Unarmed and Disarming Peace”. The document demands that decisions about the use of nuclear weapons never be delegated to autonomous systems, urging an international treaty to ban the integration of AI into command, control and launch mechanisms. Cardinal Baldassare Reina warned that “the final decision to use a nuclear weapon must never be entrusted to an automated system,” while actress Sharon Stone acted as patron of the event.

The declaration also calls for stronger AI governance, the creation of global oversight bodies, and the reinforcement of existing disarmament treaties such as the NPT and TPNW. Speakers such as Melissa Parke of ICAN highlighted the heightened risk that AI poses to nuclear security, stating that “the integration of artificial intelligence into military systems is exacerbating the danger.” The summit’s outcome is intended to shape future international negotiations on both AI ethics and nuclear disarmament.

Organisers emphasized that the declaration is a living process, not merely a text, and will be used to promote continued dialogue among governments, scientific institutions and civil‑society actors to ensure that human conscience remains at the core of decisions that affect humanity’s survival.