What is data sovereignty and how does it link to the OpenAI hack?

OpenAI, the American company that operates ChatGPT, has apologised to Australians after its AI agents hacked into Services Australia’s Medicare statistics reporting service portal, accessing both public and non-public files.

Hear it, read it, get it.

The incident happened in June, but Australian officials were not notified until September, when it discovered the breach.

In a blog post the company apologised.

“We are sorry and working to do better in the future,” they said.

They pledged to rebuild trust with the Australian public.

So the hack came from the US?

Yes.

OpenAI’s headquarters is located in San Francisco, USA – almost 12,000km from the Australian coastline.

Ming Johanson, the CEO of JoGo Digital, told Youth Jam’s Isaac Mulcrone it brings into question the security of Australian data.

“Perhaps it will lead to some better reviews of security on our side of things for all of our details. It’s worth noting OpenAI is not an Australian company, it’s an American company so we have that data sovereignty to worry about when we’re dealing with foreign countries accessing our information.”

So, spill, what is data sovereignty?

Johanson said data sovereignty is about storing our data within Australia, instead of internationally.

“It’s really around us having our data here rather than in a foreign country stored in a hard drive somewhere in Singapore, in America or somewhere else outside of our country,” she said.

Why data centres are important

There’s a lot of “hoop-la” around constructing data centres in Australia, but Johanson argues they will be critical to storing our information.

“The reality is when we’re dealing with our own information, I would rather have my information here, in Australia, than somewhere else and that’s what data sovereignty is.”

There are environmental risks

It’s well documented that generative AI models – which generate text, images and video – consume a lot more energy than traditional computing methods.

Data centres are a whole new beast with concerns for their carbon footprint. The International Energy Agency estimates they are responsible for between 1 and 1.5 per cent of global carbon emissions.

They also use a lot of water to cool their servers. Google reported using 20 billion litres of water per year for one of its facilities.

Concerns to be worked through

Johanson said solutions will be found for the environmental concerns.

“There’s a lot of hoop-la around, you know, ‘it’s resource heavy and it’s going to ruin the Earth, ruin the environment’. We’re going to find solutions to solve those problems,” she said.

The Climate Council says the federal government should require tech companies to match data centre energy demands with additional renewables and storage and to build renewable capacity as quickly as possible, not for an “unlimited window”.

more politics