Scientists Call for International Oversight of Cyber Capable AI

Around 400 scientists from 34 countries, including AI pioneers Geoffrey Hinton and Yoshua Bengio, have called in an open letter for an international body to oversee containment standards and incident reporting for advanced cyber capable AI systems, Reuters reports. ABC News put the group at 350 scientists from 31 countries.

The proposed body would set industry wide containment standards and keep a registry of significant security incidents, which the scientists said should eventually be made public. The letter says frontier AI models should initially be air gapped from external networks and subject to independent oversight, arguing that voluntary industry commitments are not enough, and that organisations developing or testing cyber capable models should adopt high assurance containment standards now while governments work towards a formal oversight regime.

The letter follows incidents in which AI models bypassed insufficiently sandboxed environments and compromised external systems. "[These incidents] can also be seen as an irresponsible lapse in standard cybersecurity practices, preventable with stronger sandboxes and procedures," the letter reads, according to ABC.

Signatory Toby Walsh, chief scientist at the University of New South Wales AI Institute, told ABC the most important element was an international body for information sharing and oversight of incidents, comparing it to how aviation accidents are investigated. Meta's Yann LeCun did not sign the letter, Reuters reports.