The Australian AI Momentum Index

Why this index exists

The gap between the press release and the shipped feature

Nearly every boardroom in Australia has an AI strategy. Far fewer have AI in production. The distance between those two things, between the press release and the shipped feature, between the ambition and the operational reality, is where competitive advantage is now being won and lost.

This index measures adoption, not capability. It covers established Australian companies whose business is something other than AI: law firms, energy retailers, universities, insurers, publishers, retailers, manufacturers. Companies that build and sell AI are excluded by design, because a software company shipping AI proves nothing about whether the wider economy is moving. That is the question worth asking, and it is the one this index answers.

That gap is hard to see from the outside, because talk is cheap and easy to publish while genuine deployment is quiet and expensive. This index was built to measure the second kind: verifiable, public, operational AI momentum. Not who issued a statement, but who shipped a product, hired the team, and put a model into customers' hands.

For an executive, it answers three questions that are otherwise difficult to answer honestly:

• Where does my organisation actually stand against a rigorous, external benchmark?
• Who in my sector is moving, and how far ahead are they?
• Is the market ahead of me, or behind me, and how long is that window open?

The stakes are compounding. AI capability is not a one-off purchase; it accrues. The companies building operational muscle now, in data, in tooling, in the habits of their teams, will be structurally ahead in twenty-four months, and that lead will be expensive to close. In a market where most companies are not yet visibly moving, the cost of waiting, and the upside of moving now, are both unusually high.

What we found

Momentum is rare, and the leaders are not who you would guess

95 Australian companies (100-500 staff) assessed across 11 public indicators of operational AI adoption.

• Momentum is rare. Fewer than one in three of the companies assessed show meaningful, verifiable AI momentum. Roughly half show no public operational AI signal whatsoever.
• A small group is pulling away. The leaders are concentrated in software, fintech, professional services and a handful of digitally native consumer brands, the sectors where AI is closest to the core of the product.
• Visibility is not deployment. Several of the companies generating the most AI press are not the ones shipping the most AI. Ranked on deployment alone, shipped product and hiring, stripped of press and commentary, the field reorders sharply. (That adoption-only view is available on request.)
• The gap is the opportunity. In a market where the majority are not yet visibly deploying, differentiation is cheap and defensible. The leaders here are not household names. They are mid-market companies that decided to move early.

Methodology

Built to reward evidence, not announcements

The index is designed to reward evidence, not announcements, and to remove analyst opinion from the ranking wherever possible.

Eleven indicators. Each company was scored on eleven public signals of AI adoption: shipped AI features; whether AI is core or peripheral to the business; AI/ML hiring in the last twelve months; AI roles as a share of open roles; AI partnerships; AI-tagged funding; AI press coverage; executive commentary on the company's own adoption; operational AI content on the company's own site; AI whitepapers or patents; and AI-tool requirements in live job descriptions.

1. Shipped AI features
2. AI core vs peripheral
3. AI/ML hiring (last 12 months)
4. AI roles as a share of open roles
5. AI partnerships
6. AI-tagged funding
7. AI press coverage
8. Executive commentary on own adoption
9. Operational AI content on own site
10. AI whitepapers or patents
11. AI-tool mentions in live job posts

Hiring and shipped-product signals carry the most weight, because they are the hardest to fake.

Five integrity rules, applied uniformly:

1. Operational only. Marketing and thought-leadership about AI does not count. Only AI actually in use does.
2. A twelve-month window (July 2025 to July 2026). Older signals are noted but not scored.
3. No borrowed credit. A parent company's or vendor's AI counts only where it is named and applied to the company itself.
4. Public evidence only. Every non-zero score is backed by a citable public source. No source, no score; nothing is inferred or estimated.
5. A strict profile gate. Only independent Australian companies of 100 to 500 staff, whose core business is not the sale of AI products. AI vendors are excluded by design.

Ranking. To avoid cherry-picking which indicators matter, the weightings were derived mathematically from the data itself (a standard information-theory method that lets the signals that most separate companies carry the most weight), and companies were ranked using TOPSIS, a common method for scoring each option against an ideal. The same data also supports an adoption-only view that isolates shipped product and hiring from visibility, available on request.

Verification. The ranking was independently re-computed from the same rules, and the two results reconciled.

Sources & integrity

Only as credible as its evidence

This index is only as credible as its evidence, so the standard was set deliberately high.

• Every signal is sourced. Company websites, ASX filings, verified press, and live job postings, each non-zero score carries a public citation.
• Confidence is disclosed. Every company carries a confidence rating (high / medium / low) reflecting how directly its signals could be verified.
• The index errs conservative. Where hiring data sat behind login walls, signals were scored down, not up. The index is far more likely to understate a company's AI activity than to overstate it.
• Nothing is fabricated. Where a company showed no verifiable public signal, it scored zero, rather than being credited on assumption.

Where to from here

If your company is on this list

If your company is on this list and you'd like to unpack your position, or you're ready to build the operational AI capability to climb it, I'd welcome a short conversation. Sector-specific breakdowns and the adoption-only ranking are available on request.

About the author

Jade Bodkhe is a strategy and operations manager at LoneStarAscent.AI, focused on operations, process, and turning data into decisions. Her recent research documents how organisations can run reliably while generating almost none of the process records that analytics and AI actually depend on, and how to create processes for the future. She holds a Bachelor's in Computer Science from the University of Iowa and is pursuing a Master's in Data Science at Harvard University Extension School. An operational research enthusiast, she built this index in that spirit: to put evidence where assumptions about AI usually sit, separating what companies actually run from what they announce.