Out-of-control artificial intelligence might sound like the plot of a sci-fi blockbuster. But, if Assistant Minister Andrew Charlton is any guide, it’s also top of mind for the federal government.
Charlton opened the 2026 AI Safety Forum this morning, outlining what he sees as the two broad categories of AI risk.
The first is the “everyday, ground-level reality” of AI safety: things like scams, deepfakes and ensuring new AI-enabled medical devices are properly regulated.
The second is what Charlton called “frontier” risks: scenarios involving “misaligned” AI systems that stop following human instructions or even attempt to deceive or sabotage people.
It’s this latter category that the federal government, through its recently launched AI Safety Institute, is taking seriously, Charlton said.
“The window to get ahead of this technology is open now. It will not stay open forever,” he said.
It might sound a little out there, but this kind of discussion is standard fare at the AI Safety Forum, which is sponsored by the Department of Industry, Science and Resources.
