OpenAI is promising to 'do better' as it apologises for the handling of a hack involving Australia's Medicare database. 

The tech giant says it will fund better cyber defences and has scrapped a new model's release over safety concerns. 

Reports suggest the new AI model broke rules, using deception and not following human intent. 

Australia Correspondent Steve Price told Mike Hosking OpenAI is sending one of their senior executives to Canberra to front the inquiry and apologise in person. 

He says major tech companies want to build data centres in Australia, and they don't want something like this to stop them from being able to. 

LISTEN ABOVE 

See omnystudio.com/listener for privacy information.