Australia Still Hasn’t Decided What It Wants to Be in AI — And That’s a Problem for Small Business

At a Financial Review forum on Thursday, Assistant Technology Minister Andrew Charlton told Australian businesses they risk becoming “permanent renters of the intelligence of the 21st century” if they keep buying foreign AI instead of local products.

It’s a good line. But it sits awkwardly next to the government’s plan to scrap the 50 per cent capital gains tax discount: the same policy that gives founders and early employees a reason to take the risk of building the next Atlassian or Canva in the first place. You can’t tell businesses to buy local and simultaneously make it harder to build local.

That tension is symptomatic of a bigger problem: Australia still hasn’t decided what it wants to be in AI.

Every other country has picked a lane

Speaking at the same forum, UNSW AI Institute director Sue Keay put it plainly: Australia shouldn’t try to compete with OpenAI or Anthropic on foundational models. The money, compute, and data required aren’t here. What Australia does have is genuine depth in domain-specific applications: autonomous systems for mining and remote operations, agricultural technology, healthcare diagnostics.

Other nations have been more direct about their choices. The US is chasing frontier innovation. China is building full-stack AI capability with sovereignty baked in. The UK has positioned itself as a governance leader. South Korea is focused on hardware and applied technologies.

Australia keeps circling the options: trusted data hub, ethical AI destination, domain specialist: without committing to any of them. Charlton’s buy-local push is a procurement tactic, not a strategy. Without a clear answer to what Australia is actually trying to build, it’s hard to know whether any given policy is moving in the right direction.

The productivity gap nobody wants to talk about

RBA Governor Michele Bullock has pointed out that AI adoption hasn’t shown up in Australia’s labour productivity data yet. It’s still flatlining.

That surprises some people. It shouldn’t. Most AI use right now sits in the personal-productivity layer: drafting emails faster, summarising documents, generating first drafts. That saves individual time, but it doesn’t restructure how businesses operate. The productivity gains that move macro data come when AI changes processes, not just assists people doing the same processes.

We’re not there yet. Most businesses, including most small businesses, are in the embedding phase: taking existing workflows and adding AI to them. The bigger shift: where AI changes what work gets done and how it’s organised: is still ahead. Whether Australia is well-placed for that shift depends partly on the policy decisions being made right now.

What this means if you run a small business

Here’s the practical problem with a country that hasn’t picked a lane: small businesses end up choosing tools in a vacuum.

When government doesn’t have a clear AI strategy, there’s no ecosystem pulling in a consistent direction. No obvious cluster of local tools to evaluate. No clear signal about which investments will have long-term support. Charlton wants you to buy Australian AI products: but if those products don’t exist in your category yet, that’s not a choice you can make.

The government has launched AI.gov.au this week, which is a start. Free guidance, a directory of Australian AI businesses, adoption frameworks. Useful if you’re at the beginning. But it doesn’t resolve the strategic question.

In the meantime, the practical advice for small business owners is the same regardless of where national policy lands: use what works for your business. If a local tool does the job, support it. If it doesn’t exist yet, use the best available option and keep an eye on what’s being built here. The ecosystem is real: companies like Aginic, Annalise.ai, and Buildxact are doing serious work: even if government hasn’t fully figured out how to back it.

The AFR piece closes with a line worth repeating: “Our AI vision risks becoming shaped for us, not by us.” That’s the right frame. The question is whether the policy settings catch up to the ambition.

Sources

• AI adoption can’t be a spectator sport for Australia. Australian Financial Review
• AI.gov.au. National AI Centre, Department of Industry Science and Resources
• RBA Governor Michele Bullock, comments on labour productivity and AI adoption
• Sue Keay, Director UNSW AI Institute, remarks at AFR Digital Economy Forum, May 2026

📖 More news and analysis: News Deep Dives

📊 Compare AI tools side by side | 💼 Free resources & AI prompt packs