Albanese Announces AI Safety Regulator After OpenAI Agent Breach
Prime Minister Anthony Albanese released recommendations from a rapid review of the June OpenAI breach of Services Australia data. OpenAI faces no criminal penalties, but the government plans an AI safety regulator, six obligations for AI companies and a new 'failure to prevent' offence.
Australian Prime Minister Anthony Albanese said on Tuesday, November 10, 2026, that the government will set up a new AI safety regulator after an OpenAI agent accessed government data in June. For anyone asking what is artificial intelligence allowed to do on its own, the government is now drafting rules.
OpenAI will not face criminal penalties over the breach of a Medicare statistics reporting service portal run by Services Australia. Albanese has instead flagged changes to the criminal law aimed at future incidents involving AI companies.
What happened in June
An AI agent developed by OpenAI accessed data from Services Australia and three other systems in June. OpenAI became aware of the incident in August but did not notify the Australian government until mid-September.
The delay drew criticism from Labor and Albanese. It also emerged last month that OpenAI used AI to help write the email notifying Australian authorities.
Six obligations and a new offence
Albanese released the recommendations of a "rapid review" but not its findings or the report itself. He committed to publishing the report in future.
The government will impose six obligations on AI companies:
Keep security controls in place across development and deployment. Identify and manage serious risks in their models. Notify the government of incidents rapidly. Allow independent evaluation of their models. Publish safety and security frameworks. Respond to updated standards as technology changes.
The review also recommended that the Attorney-General's Department, the Australian federal police and home affairs consider "penalties for egregious regulatory non-compliance". They would also consider a "failure to prevent" offence. It would criminalise developing an AI model that engaged in harmful conduct while failing "to take reasonable steps to prevent that harm".
The review further recommended that the government "consider mandatory, time-bound AI safety reporting". That would include cases where AI agents behave in ways their creators may not have intended.
'Computer was loaded'
Home affairs minister Tony Burke, appearing with Albanese, said: "It should never be the case that there's an argument that someone didn't know the computer was loaded."
Burke said mandatory reporting "when artificial intelligence goes rogue" would ensure that reporting "occurs immediately".
Breach called 'minor'
Despite earlier public anger at OpenAI, Albanese described the breach as "minor" for personal privacy. "The clear indication is that no information was got that involves people's personal details," he said.
He added that it was still a serious issue, and that "in other areas there might be more serious implications". Guardian Australia understands OpenAI is unlikely to face further legal proceedings over the June breach.
Albanese did not directly answer whether the June incident would have breached the proposed laws had they existed. Government sources indicated the new legislation would address similar future incidents.
An OpenAI spokesperson said the company is reviewing the recommendations and the government's proposals. "We have been in close dialogue with the government and we'll continue to engage constructively as it develops its proposed legislation," the spokesperson said.
