Subtitled "Toward an Unarmed and Disarming Peace," adopted by participants of the Global Nobel Laureates Assembly at Borgo Laudato Si', Vatican, July 2026, and published July 16. Signatories include Nobel Peace laureates Juan Manuel Santos, former President of Colombia and chair of The Elders, and Denis Mukwege.
Text captured from The Elders' mirror; the canonical assembly page is JavaScript-rendered.
The framing
The declaration's organizing move is to treat the two technologies as one problem rather than two: "More than eighty years after the dawn of the nuclear age, and at the threshold of the age of artificial intelligence, we are presented with an unprecedented challenge."
Its historical argument is that the nuclear precedent is a record of a missed opportunity rather than of successful management: "When nuclear weapons were first being developed, the world had a chance to avoid living in a permanent dynamic of fear. We failed, and as a result, humanity has lived at the edge of nuclear annihilation."
Four interaction channels are named by which AI raises nuclear risk, and only the first concerns AI inside nuclear systems:
- Compressed decision time. "AI built into nuclear systems leaves little time for, or even replaces, human judgement in a crisis."
- Economic competition. "AI will likely cause mass job loss and sharpen economic competition between nuclear-armed states."
- Cyber. "AI uplifts cyber attacks that can fracture critical civilian and strategic infrastructure."
- Information warfare, with the chain stated explicitly: "Peace depends on a shared reality. Without facts there can be no truth; without truth there can be no trust; and without trust, no dialogue, no verification, and no restraint among nations."
Concentration is identified as a structural aggravator: "The most advanced AI capabilities, computing resources, and data infrastructures are becoming increasingly concentrated in a small number of countries and corporations, creating significant asymmetries of power and limited incentives for cooperation."
The declaration invokes Pope Leo XIV's call toward "an unarmed and disarming peace," rejecting "the belief that lasting security can be built on fear, domination, or the permanent threat of mutual destruction," and closes with Russell and Einstein: "We appeal as human beings to human beings: remember your humanity, and forget the rest."
The six principles
1. Disarming the next arms race. AI is treated as "an arena of strategic competition, like nuclear weapons before it," with a call for action "through governance arrangements such as benefit-sharing mechanisms, the promotion of AI applications that serve human wellbeing, and restraint in its most destabilizing applications… while the choice remains ours."
2. Responsible development. Two demands, the second the most specific technical commitment in the text. Developers "should transparently publish the principles that guide the behavior of their models and be held responsible and liable for those models' adherence to such principles" — converting published specifications into a liability standard. And:
"no organisation should initiate, and no government should permit, fully-automated recursive self-improvement in artificial intelligence systems without the means to monitor, and if needed, to halt such systems."
The condition is capability-based rather than prohibitory: what is barred is proceeding without monitoring and halting capability, not the activity itself.
3. Responsible use. The bright line: "An automated system should never make the final decision to launch a nuclear weapon," with a call for "an international treaty prohibiting the reckless integration of artificial intelligence into nuclear command, control, and launch systems so that meaningful human control is maintained." Nuclear states are asked to "engage in fail-safe reviews to reduce the vulnerability of their nuclear forces to unauthorized control or manipulation by AI." The principle also affirms beneficial use "to improve human well-being, accelerate scientific and medical progress, protect the environment, strengthen resilience, and advance peace."
4. Responsible governance. A call on governments, corporations, and international organizations "to enable coordinated slowdown of frontier AI development by establishing shared mechanisms, such as verification and robust internal and external evaluations," alongside support for the UN Independent International Scientific Panel on AI and for "expanded research and dialogue to propose new institutional pathways for international governance."
5. Responsible leadership. Focused on the next generation — education, mentorship, and "intergenerational partnership," advancing "ethical responsibility, critical thinking, media and AI literacy, and a culture of peace." It also notes that connected threats' "unintended consequences… often impact those who do not have access or control over such technologies," and calls for "a digital commons which grows data to advance understanding and action."
6. Nuclear disarmament. Calls for "urgent, sustained, and good-faith negotiations leading, within an agreed and time-bound framework, to the verifiable and irreversible elimination of nuclear weapons," rejecting "the normalization of nuclear weapons as permanent instruments of security" and invoking the NPT and the TPNW. The operational asks to nuclear-armed states are the standard risk-reduction set, several of which bear directly on the AI-compressed-decision-time concern: "lower alert levels; extend decision time; move away from launch-on-warning and prompt-retaliatory postures; reduce the role of nuclear weapons in security doctrines; strengthen crisis communications; prevent cyber operations against nuclear command, control, and communications."
Relationships
- supports: Religious and Civil-Society Voices on AI — the Vatican-hosted assembly's adopted text
- related: When AI Builds Itself (The Anthropic Institute, June 4, 2026) — the recursive-self-improvement condition parallels its verifiable-slowdown position
- related: The Elders — signatory organisation and text mirror
- related: Recursive Self-Improvement (RSI), Autonomous Weapons, Preliminary Report of the Independent International Scientific Panel on AI (United Nations, July 2026), AI Existential Risk