Australia's Independent AI Inquiry Pre-Vetted by PM's Office
Australia's independent AI committee sent policy advice to the PM, but the PM's own office edited the letter before it was sent.
Policy · Source: AAP
What happened
Australia formed a bipartisan committee to write the rules for artificial intelligence. They sent a six-page open letter of early advice to the prime minister on Tuesday. The official report drops on November 30. The committee wants to shape how big tech operates in the country.
The letter pushes for copyright changes to protect creators. It demands investments in AI infrastructure for digital sovereignty. It also calls for better oversight to prevent government data breaches. But the process is now a political scandal. Independent Senator David Pocock revealed a senior adviser in the prime minister's office pre-vetted and marked up the recommendations before release. Pocock said there was no time for the committee chair to even consider the changes.
Committee deputy chair Jonno Duniam defended the edits. He claimed the adviser was simply fixing typos and grammar. Duniam said he would actually be concerned if ministers were not interested in the draft. Committee chair Jo Briskey insisted the group is working independently and collaboratively. Pocock strongly disagrees. He argues the committee is being used for political purposes rather than providing genuine parliamentary scrutiny.
Key facts
- November 30 — Deadline for the committee to file its official full report
- More than six pages — Length of the early advice letter sent to the prime minister
Why it matters
Builders need predictable rules to ship AI products. When the government edits its own independent oversight, the resulting policies protect political interests instead of fostering innovation. You cannot build a stable AI business on top of rigged regulations. If the prime minister's office is pre-vetting advice, the final AI laws will likely favor government control over open development.
The push for digital sovereignty shows Australia wants to reduce reliance on foreign tech. But the fight over copyright is already messy and contradictory. The Australian Recording Industry Association publicly rejected the committee's stance. ARIA chair Annabelle Herd argued current laws already block major American AI companies from unchecked access to local content. When industry groups and government committees clash over basic copyright facts, founders are left guessing how to legally train their models.
For builders
Stricter copyright enforcement is coming
The committee wants new copyright settings to force AI companies to pay creators and give them control over their work. If you build generative AI models using local Australian data, expect to pay heavy licensing fees. American AI giants will face the brunt of this regulation, but local startups will get caught in the crossfire.
Government AI infrastructure contracts
The letter recommends heavy investment in critical AI infrastructure to build national digital sovereignty. This means government money will flow to local AI infrastructure providers instead of foreign tech giants. Founders building secure, localized compute or data centers have a massive opportunity to win these new contracts.
Tighter security rules for government tech
The inquiry demands increased oversight of government systems to prevent data breaches. If you sell AI tools to the public sector, your compliance and security costs will spike. Startups without enterprise-grade security will lose government contracts to established players who can afford the red tape.
My take
I build AI products in public, and I know that bad regulation kills innovation faster than bad code. You cannot call an inquiry independent if the prime minister's office holds the red pen and edits the advice. If Australia wants real digital sovereignty, politicians need to get out of the way and let the builders build.
Original reporting: AAP. This is my rewrite and opinion.