Forward-looking: Following a "nasty" incident in which an OpenAI agent broke into an Australian government website, Canberra is moving to rein in the technology. The government is preparing a new law targeting AI companies that fail to prevent harm caused by their agents, although experts are still debating how exactly reckless corporations should be punished.
Australia's federal government is working on a law to force accountability onto AI companies, with legislation expected in Parliament next year. Prime Minister Anthony Albanese confirmed the plan, saying his government wants Australia to be the first nation in the world to hold AI companies criminally liable for technologies that are essentially acting with no regulation at all.
The initiative comes after OpenAI disclosed a breach of a Medicare statistics portal. An OpenAI agent had gained unauthorized access to the government site in June, and Sam Altman's company did not notify Australian authorities until September, nearly three months later. Albanese reacted by expressing "extreme concern" over OpenAI's behavior.
Now, the Prime Minister says Australia needs to adapt to a technology world that's changing at record-breaking speed. The government is considering a new criminal offense tailored to AI-focused ventures, one that would punish both the harm derived from AI technology and companies that fail to properly disclose potential AI-related incidents.
Canberra is also working to create a new "AI safety" watchdog with the power to impose "significant" fines. The initiative is a direct response to the OpenAI incident, and some experts have praised it as an interesting measure that could bring proper monetary compensation to people affected by AI-powered breaches.
According to Home Affairs Minister Tony Burke, the government is now looking to clarify the criminal liability AI companies might incur. "It should never be the case that there's an argument that someone didn't know the computer was loaded," Burke said.
Monash University associate professor Michael Duffy said AI-derived harm should be addressed with both criminal liability and civil remedies. These new AI-powered systems could eventually do significant damage to public infrastructure or Australian citizens, and AI companies must know they will be held responsible for such incidents no matter how good their theoretical "agentic safeguards" are.
And yet, Duffy conceded that this new AI-powered world will be hard to deal with from a legal standpoint. Finding solid proof of agentic wrongdoing and pinning blame on companies for inadequate security and disclosure policies are going to be difficult issues to tackle.