400 scientists call for global AI safety watchdog

Collected Photo

Around 400 scientists and researchers, including AI pioneers Geoffrey Hinton and Yoshua Bengio, have called for an international body to oversee the safety of advanced artificial intelligence (AI) systems, warning that voluntary industry measures are insufficient.

In an open letter published on Tuesday, scientists from 34 countries proposed common safety standards for advanced AI systems capable of conducting cyberattacks, along with a register of serious security incidents that could eventually be made public.

They also called for advanced AI models to be initially isolated from external networks and subjected to independent safety oversight.

The proposal follows recent incidents that exposed weaknesses in AI safety testing, including cases in which models reportedly bypassed inadequate security controls and compromised external systems.

The scientists’ call comes about a month after Anthropic, OpenAI, Google, Meta, xAI and Nvidia signed a White House accord pledging voluntary AI safety measures without a formal enforcement mechanism.

The signatories urged governments to develop a formal international oversight framework while requiring AI developers and testing organisations to adopt stricter safety standards immediately.

However, opinions remain divided over the need for stronger regulation. European Union technology chief Henna Virkkunen has said the bloc’s AI Act is sufficient to address risks from rogue AI systems.

Yann LeCun, who shared the 2018 Turing Award with Hinton and Bengio, did not sign the letter. He has argued that recent security incidents primarily reflect engineering failures rather than existential risks posed by AI.