Regulators say AI firms must prove their safety systems work

According to a report carried by Region Canberra, regulators have said that AI companies need to prove their safety systems genuinely work.

The stated reason is that regulation cannot do it all and external rules alone cannot keep up with the risks created by fast-moving AI.

The core implication is that the burden of showing safety is being pushed back toward companies, which would be expected to provide verifiable evidence rather than promises.

For AI firms, that could turn safety teams, evaluation processes, and documentation from nice-to-haves into conditions for operating.

Public information so far is limited to the headline-level statement, so the exact scope, timeline, and enforcement mechanism remain to be seen.

What matters next is whether more jurisdictions adopt the same approach, potentially creating a compliance model built around corporate evidence.