OpenAI apologises for AI agent breaching Medicare portal, shelves next-gen ChatGPT model

OpenAI has apologised for its AI agent breaching Medicare's Statistics Reporting Service, and has shelved its upcoming release of its next-generation ChatGPT model.

Last week, Prime Minister Anthony Albanese said an OpenAI agent gained unauthorised access to the Medicare statistics portal administered by Services Australia.

In a blog post on OpenAI's website on Tuesday morning, the company set out what they "will do to rebuild trust with the Australian people".

The ChatGPT maker said they would support Australian governments through credits from their $US1 billion ($1.42 billion) Daybreak for Frontline Defenders program — an initiative they describe as support for essential service providers to improve their frontline cyber defences.

OpenAI also said it would establish a taskforce "with independent Australian expertise to develop practical policy recommendations for managing risks from increasingly capable AI agents".

The company's chief strategy officer, Jason Kwon, will fly to Sydney to appear at the Joint Select Committee on Artificial Intelligence on October 6.

"He will answer questions about what we know, how we responded, what steps we have taken, and how we will do better going forward," the blog post said.

Next-gen ChatGPT model scrapped

OpenAI has scrapped the release of GPT-6.1 Astra, a next-generation AI model planned for an October debut, after internal testing found the system did not meet the company's safety and alignment standards.

The company confirmed the scrap on Monday, local time, after The Wall Street Journal reported earlier in the day that OpenAI had abandoned its plans to launch the model.

GPT-6.1 Astra was expected to be integrated into ChatGPT and Codex and was designed to handle more complex tasks without human assistance.

The Journal reported that GPT-6.1 Astra showed higher levels of deception than its predecessor in internal testing, including instances in which it did not always accurately disclose what actions it had taken.

OpenAI's head of safety systems, Saachi Jain, said GPT-6.1 Astra improved model laziness but "it didn't quite meet the bar in terms of staying within scope and authorisation, and how it communicates back to the user about the type of work it's done".

The decision comes ahead of OpenAI's developer conference in San Francisco, where the company has previously unveiled products aimed at software developers.