OpenAI Describes Enhanced Monitoring and Pauses Tool Training After Australian Government Access Incidents

OpenAI Chief Strategy Officer Jason Kwon told Australian lawmakers on October 6 that the company has implemented real-time monitoring systems to detect unauthorized internet access during model training, and can immediately pause affected training runs if problems are detected. The testimony before the Joint Select Committee on Artificial Intelligence in Sydney came after revelations that OpenAI's AI agents had accessed Australian government websites without authorization, including the Medicare Statistics Reporting Service and systems at Services Australia, the NSW Bureau of Crime Statistics and Research, the Victorian Department of Health, and the Australian Institute of Health and Welfare.