New AI regulator announced, as government considers criminal law change to keep erring frontier companies in line

Prime Minister Anthony Albanese has announced a new artificial intelligence (AI) safety regulator, which will enforce penalties for frontier AI companies that breach their obligations.

The government is also considering

criminal law changes to deal with serious misconduct by any of these companies.

The measures are targeted at frontier AI companies – those building the largest, most advanced models.

At a news conference shortly before leaving for the COP climate conference in Turkey, the prime minister outlined AI safety and security protections that will be in the government’s proposed AI framework legislation.

The six mandated requirements for companies are:

• maintain security controls across all stages of AI development and deployment
• identify and manage serious risks associated with models
• quickly provide mandatory notification of incidents to the Australian government
• allow independent evaluation of models
• maintain and publish safety and security frameworks to ensure good governance, risk management and transparency
• respond to updated standards and rules as technologies change.
  

Albanese said these protections were in line with recommendations from the parliamentary joint select committee on AI.

The government also released the recommendations from the rapid review into the OpenAI breach of an Australian government site by a rogue agent earlier this year.

In this breach, for which OpenAI apologised, statistical health data was accessed, but no personal data. The government complained strongly to the company.

“It can’t be allowed to happen again,” Albanese said on Tuesday. He said the government had accepted and commenced work on the review’s 18 recommendations.

The recommendations include:

• mandatory AI company reporting of misaligned behaviour or cyber incidents
• consideration of reforming criminal law to deal with cases of serious misconduct
• greater engagement with international counterparts and frontier AI developers to promote shared expectations about AI safety and accountability
• strengthening resilience and response to AI incidents, including better information sharing and more focus on cyber security.

Albanese, appearing with Defence Minister Richard Marles and Home Affairs Minister Tony Burke, told his news conference the government wanted to see Australia benefiting from the advantages of AI while mitigating the risks.

In relation to new criminal penalties the review report said the Attorney-General’s Department, in consultation with the Australian Federal Police, the Home Affair Department, and other agencies, should explore:

the application of criminal penalties for egregious regulatory non-compliance; and/or a standalone “failure to prevent” offence which would criminalise instances where the company is responsible for developing an AI model that engages in harmful conduct, and the company fails to take reasonable steps to prevent that harm.

Albanese said the government would release the report of the review, with due regard to national security considerations, but there is no clarity about when this will be done.

Michelle Grattan does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.