Why a $3 Million AI Royal Commission Is Solving the Wrong Problem

South Australia announced a $3 million Royal Commission into AI, timelined at nine months. Paloma Newton, founder of the pet-longevity platform Elita, doesn't object to government taking the technology seriously. She objects to the vehicle. "Nine months in AI time is probably three lifetime iterations," she says. "It's like using a hammer to make a pancake."

What sharpens her frustration into something more pointed is the timing: the state's premier signed a memorandum of understanding with OpenAI in the US, then flew home and announced an inquiry into the risks of the exact technology he'd just partnered with. "That just kind of peed me off," she says. "From a government perspective, they need to be a little bit impartial."

A faster, less theatrical alternative

Her counter-proposal isn't a bigger inquiry — it's a standing one. Rather than a single nine-month commission that's out of date the day it publishes, she'd build a revolving advisory board stocked with the people actually building and living with the technology: founders like Tom Kelly at Heidi Health and Melanie Perkins at Canva, alongside teachers, lawyers and ethicists, refreshed on an ongoing basis rather than convened once. "That's kind of what you need to do," she says, "when the technology is constantly evolving."

What we're actually afraid of

Newton's sharper claim is that the public anxiety around AI is misdirected. Listening to a podcast about data centre backlash the morning of this conversation, she landed on a theory: towns don't organise against data centres because of the buildings themselves — the same physical footprint, cooling systems and internet infrastructure have existed since the early internet with no equivalent backlash. What's different now is trust. Many of the same communities have already watched mining companies buy land, promise jobs, and leave a county millions of dollars in the hole with nothing built. A data centre becomes the tangible thing to be angry at, standing in for something harder to name: that a handful of companies, mostly American, are steering the biggest technological shift of most people's lifetimes, inside institutions that have already let people down before.

"I don't think we're scared of data centers. I don't even think we're scared of AI. I think we're scared of the fact that three people are going to own so much of this development — and we have no trust in the systems that are built to protect us."

Her response to the reflexive outrage this produces — the framing that using AI at all makes you a bad person — is that it's reductive and, worse, ineffective. "Putting everything on the individual is a race to the bottom," she says. Pointing at any one person's tool use avoids the harder, more useful argument: pushing for better systems underneath the technology, like closed-loop water cooling and renewable-powered data centres, rather than a wholesale rejection of the category.

Efficiency was never going to buy anyone free time

Newton is equally blunt about a separate piece of AI mythology: the idea that automating your work frees up your time. She points to the Jevons Paradox — the nineteenth-century observation that making steam engines more fuel-efficient didn't reduce coal consumption, it increased it, because efficiency just uncovered more uses for the resource. "We are incredibly output-oriented," she says. "I just never thought this was going to make us work less." Her own discipline against that pull is structural: she reviews her diary at the start and end of every week, checking whether her time actually moved toward the company's strategic goals or was spent "outputting for the sake of feeling in control" — her honest description of her own anxious default setting.

Built backwards, on purpose

Elita's own origin story runs against the same instinct the Royal Commission critique targets: moving fast for the sake of moving fast. When Newton adopted her dog Edgar — a golden retriever cocker spaniel predisposed to arthritis and hip and elbow dysplasia — she went deep into breed-specific health data most owners never see. Rather than launch a product and market it, Elita built a stem cell bank first, then spent years developing the underlying science before its first marketing push, with a team of vets and biologists and no commercial hire at all until the week of this recording. "I would feel some kind of way if we overpromised and then underdelivered," she says. "The grief of the death of a pet is something we haven't quite understood in society yet. Anything I can do to slow that down, I will."

The cost of half the room opting out

Newton's final concern is more structural than technical. Women, she argues, disproportionately carry the ethical weight of new technology — the same pattern she's observed around fast fashion — and are correspondingly more likely to step back from AI over legitimate ethical discomfort. Her worry is what gets built if that happens: she points to Verity AI, built to detect counterfeit fashion; Ovum, built around a gap in women's health research including endometriosis; and Andromeda's aged-care robotics, tackling a category few well-funded technologists have prioritised. "These kinds of products will not be developed by men," she says, "because it's potentially not interesting to them." The risk of opting out on principle, in her framing, isn't just personal. It's that the specific problems only certain builders think to solve simply don't get solved at all.