Government agencies fail at the first hurdle in AI self-reporting policy

Dozens of government agencies have failed the first test for overseeing their own use of AI after Australia moved away from stricter European-style AI rules and instead relied on existing regulators and agencies.

Last year, the federal government opted for a gentler model for regulating artificial intelligence than the European approach of a uniform AI law with binding rules for high-risk AI.

The federal government said existing laws would be sufficient, but tasked each agency with overseeing its own use of AI and then regulating the use of the new technology in its own industries.

But documents filed in the Senate by the government’s technology agency, the Digital Transformation Agency (DTA), highlighted the challenge of these approaches.

More than half of the authorities reviewed by the DTA did not meet the first transparency deadline for disclosing their use of artificial intelligence, according to documents tabled in Parliament.