Despite consensus among tech leaders on the need for AI guardrails, global powers remain sharply divided on how to oversee the technology
In an unprecedented briefing at the United Nations Security Council in New York, top executives from leading artificial intelligence firms warned world leaders that rapidly advancing AI technology could pose an existential threat to all humanity if left unchecked. Addressing the 15-member council during the annual UN General Assembly gathering, industry leaders called for urgent global safeguards to prevent systems from slipping beyond human control.
The landmark session saw Anthropic Chief Executive Dario Amodei warn that, if managed poorly, AI could become a risk to humanity as a whole. Echoing these concerns, OpenAI CEO Sam Altman cautioned that the world risks losing control of its future to autonomous algorithms, emphasising that critical decisions regarding AI governance must be shaped through democratic processes rather than inside technology laboratories alone.
Runaway Algorithms and Autonomous Risks
The briefing underscored growing anxiety among computer scientists and international diplomats that increasingly capable AI models could soon self-improve, operate autonomously, and threaten global security. AI pioneer Yoshua Bengio, co-chair of the UN Independent International Scientific Panel, told the council that the dangers are real and imminent.
He warned that member states face a severe, unprecedented threat that no single nation can contain alone, highlighting risks ranging from automated cyber warfare to algorithms interacting unpredictably to trigger unintended military conflict. UN Secretary-General António Guterres reinforced these warnings, reiterating calls for strict guardrails and a global ban on lethal autonomous weapons operating without human oversight.
