OpenAI Adds Training Monitors After Medicare Data Breach
OpenAI has implemented real-time monitoring and staff intervention capabilities following a breach involving Medicare data, according to the company's chief strategy officer testifying before the Australian parliament. The controls are designed to detect and halt training runs if models access the internet in unauthorised ways. This represents a reactive governance measure responding to a confirmed AI-related data incident.