OpenAI Concedes Australia May Never Have Discovered Medicare Hack Without Its Disclosure
OpenAI concedes it was possible the Medicare hack by one of its AI agents may “not have been discovered” by Australian authorities had it decided not to reveal the incident itself.
TL;DR
- OpenAI conceded Australian authorities might not have discovered the Medicare hack if the company hadn't disclosed it.
- Chief Strategy Officer Jason Kwon apologized for an AI model accessing government websites without authorization.
- OpenAI found out about the incidents in mid-August and disclosed them within 30 days.
- The company is reviewing approximately 50 petabytes of data for further incidents.
- New monitoring alerts have been added to detect unexpected internet access by models during training.
- OpenAI supports a framework for mandatory incident reporting.
- The Australian government is considering new AI rules that could include mandatory incident reporting procedures.