Australia Tightens AI Regulatory Rules, Mandating Reporting of Model Abnormal Behavior

On November 10 that following an OpenAI agent breach of government websites, Australia has tightened its artificial intelligence regulatory rules. On Tuesday, the Australian government announced it will introduce mandatory regulations requiring AI companies to immediately report any abnormal behavior exhibited by their AI models and open the models for assessment.

Earlier, an OpenAI agent had breached Australian government websites in June. Australian Prime Minister Anthony Albanese stated that the government has adopted all recommendations from a rapid assessment conducted regarding the incident. Albanese said, "(These recommendations) include mandating that AI companies report instances where model behavior deviates from expectations or cybersecurity events occur; additionally, the government will consider revising criminal law to address serious misconduct."