European lawmakers, Nobel laureates, and former heads of state demanded binding international limits on harmful AI.
They launched the initiative Monday at the UN’s 80th General Assembly in New York.
The campaign urged governments to agree by 2026 on “red lines” prohibiting dangerous AI uses.
Signatories included Mary Robinson, Enrico Letta, several MEPs, ten Nobel laureates, and leaders from OpenAI and Google.
They warned of threats like engineered pandemics, disinformation, rights abuses, and loss of human control.
Over 200 prominent figures and 70 organisations from politics, science, and industry endorsed the appeal.
Growing Concerns Over AI’s Real-World Harms
A psychiatric study revealed chatbots gave inconsistent or unsafe responses about suicide.
Researchers warned such flaws could worsen mental health crises and cause deaths.
Nobel Peace Prize winner Maria Ressa said unchecked AI could drive “epistemic chaos” and systemic abuses.
Yoshua Bengio stressed societies cannot manage risks from rapid model development.
Advocates noted past treaties banning nuclear weapons and human cloning as models for AI rules.
They said EU legislation helps but cannot regulate a borderless technology alone.
Push for Global Treaty by 2026
Supporters demanded an independent body to enforce safeguards against irreversible harm.
Ahmet Üzümcü warned that inaction could cause “irreversible damages to humanity.”
They suggested prohibitions on AI launching nuclear strikes, conducting mass surveillance, or impersonating humans.
The campaign argued only a worldwide agreement can guarantee compliance and safety.
Supporters urged quick negotiations so the UN can adopt a resolution by 2026.
