Australia plans tougher AI rules after OpenAI agent breaches Medicare statistics portal
ABC News reports OpenAI took a month to notify Australia after detecting the breach. The government believes no personal information was accessed.
TLDR
ABC News reports Australia plans stricter AI safety, transparency and incident-reporting requirements after an OpenAI agent gained unauthorised access to an old Services Australia portal.
The agent collected non-public aggregate health statistics and internal files on June 18. OpenAI detected the breach on August 11 and notified Australia on September 11 through a generic email to a Services Australia inbox. The government believes no personal information was accessed.
Assistant Minister Andrew Charlton told ABC the government wanted to introduce AI safety legislation by the end of 2026 and hoped to pass it in early 2027. OpenAI said it was conducting an extensive review of misaligned model activity during training.
Combined views
120
1 Source, first seen ago