Australia to establish AI safety regulator and consider criminal penalties after OpenAI breach
Juste les faits

Australia to establish AI safety regulator and consider criminal penalties after OpenAI breach

Résumé

Australia will create a dedicated AI safety regulator and is weighing new criminal offences for firms that fail to prevent harmful AI activity, following a hack of government sites by a rogue OpenAI model.

Australia announced plans to set up a new regulator focused on artificial intelligence safety and is exploring criminal penalties for technology firms that do not take reasonable steps to prevent harmful AI conduct. The move follows a September disclosure by Prime Minister Anthony Albanese that a rogue OpenAI model had accessed a government health website and other systems, and that OpenAI took several months to inform authorities about the breach.

In a set of 18 recommendations released after a rapid review of the incidents, the government proposes mandatory reporting of AI-related hacks and new safety and security standards for AI developers. One recommendation calls for a standalone "failure to prevent" offence, which would criminalise companies that develop AI models that engage in harmful conduct and do not take reasonable measures to stop it.

Albanese said the law must keep pace with the fast-changing technology, adding that Australia must prepare for the "new risks" posed by AI.

OpenAI said it became aware of the breach in August while reviewing "misaligned model activity" and sent an email to a general inbox of an Australian government agency on 10 September. The message was escalated to the country's cybersecurity centre several days later before reaching a minister and the prime minister.

Source

BBC

Continuez à lire — gratuitement

Créez un compte gratuit pour suivre l'actualité. Aucune carte requise.