[Photo Credit: AAP Image/Rounak Amini]

Australia needs to slow down its artificial intelligence use and create better security laws, the companies creating the technology have told the federal government.

In a submission to an ongoing inquiry, OpenAI said national leaders had a responsibility to work with other countries to stop global security threats.

It said Australia was in a narrow “defender’s window” during which it could use the technology to find and fix cybersecurity threats, rather than create them.

Tech giant Atlassian, which runs AI system Rovo, used its submission to urge the government to bolster its laws.

International standards and work already undertaken overseas should be used to protect Australian consumers, it said.

The comments come after OpenAI’s chief executive Sam Altman and Elon Musk displayed rare unity in backing Anthropic’s head Dario Amodei, who called for a slowdown in “reckless” AI development through a three-part essay.

Several cyber hacks have been linked to large AI developers that have been unable to contain their models. While the companies have privacy and security frameworks, they have regularly been criticised as doing too little.

Government departments will front the inquiry on Friday and are likely to be probed on their security measures.

One person set to appear, Australian Signals Directorate director-general Abigail Bradshaw, warned earlier this week that intelligence and security agencies needed to work with AI companies to identify early risks.

Australia’s outdated technology was little match for potential hacking as machines grew smarter, which she said needed to be addressed proactively.

Despite that, the intelligence agency has followed dozens of other government departments in beginning to use AI to streamline efficiency.