OpenAI breached Australian government websites, Greens senator says

OpenAI breached Australian government websites and took information it was not entitled to, Greens Senator Sarah Hanson-Young told ABC News Breakfast.

Hanson-Young said the breach was particularly striking because it occurred while OpenAI executives were in Canberra meeting with government officials to seek weaker copyright laws and fast-track approval for data centre infrastructure. She called the behaviour inconsistent with the company's requests for Australian cooperation.

"The fact that they have breached Australian government websites, they do need to be held accountable for this," Hanson-Young said. "And I do think that the government needs to investigate and work out exactly how it happened, how they can avoid it happening again."

The government has established a task force to investigate whether a criminal act occurred and is working closely with OpenAI to understand the steps taken by the company's AI agents during the incident, according to government statements cited during the interview.

Hanson-Young said the breach demonstrated why artificial intelligence does not yet have social licence in Australia. "AI does not have a social licence in this country and behaviour like this really shows why," she said. "And it's gonna be harder and harder for AI to earn a social licence if they behave like this."

OpenAI has said it will continue to notify of any other unintentional breaches. When asked whether it was better to work cooperatively with companies willing to be transparent and invest in the country, Hanson-Young rejected that framing.

"Of course you need companies that are willing to work with you, but you need companies who are willing to work within the law," she said. "This is the point here. Everybody else has to abide by the law. All other companies have to abide by proper regulation. AI should be no different."

Inquiry into data centres

A Greens-led Senate inquiry is examining the rollout of data centres across Australia. The inquiry has called for the heads of the two major AI companies, Sam Altman and Dario Amodei, to appear before it. Public hearings began last week in Sydney and Melbourne.

Hanson-Young said residents in Sydney, Queensland, Victoria, Tasmania, the Northern Territory and South Australia felt railroaded by the speed at which big AI companies wanted to establish operations. "We started our public hearings last week," she said. "Residents are really worried, Australians are concerned and experts are saying, hang on a minute, we just don't have the rules and regulations in place before rolling out the red carpet."

She said the inquiry hopes representatives from Anthropic and OpenAI will appear at hearings this coming week. "I think it would be very bad form if they refuse to show up," Hanson-Young said. "They want Australians to give over their land, their water, their energy, and even their own data and intellectual property. The least they can do as these big AI companies is to front the Senate inquiry."

Hanson-Young outlined concerns residents and experts have raised about the data centre rollout, including large water consumption, substantial energy demands that could push up energy prices, increased pollution and potential job losses. She also said the companies want to weaken copyright laws and are attempting to access copyrighted material.

"We need proper investment in our AI security agencies and we need to make sure that we're not just rolling over because we're starry-eyed with the dollars that they're promising," Hanson-Young said. "Australians are rightly sceptical."

She said that because of the power of the technology and serious warning signs, proper rules and regulations are essential. "Everybody else has to stick by rules. Why shouldn't big AI and the trillionaires?"