Best AI Influencers in Australia (2026): Pick by Kind of Voice

Choose the kind of voice before the name. Researchers and ethicists tell you where AI is heading and what it should and shouldn't be trusted with, founders and operators show you a business running on it, and tool testers save you hours of trial and error. Which one helps depends on whether you want rigour, a worked example or a quick answer to which tool does which job.

Everything I say about the eight people below sticks to their own page as of Nov 2026 or, in one case, to what Wikipedia states. There are no positions, follower counts or awards. I've put myself on the page as a practitioner, so read my part as the least neutral.

Stages and inboxes, not feeds

Here's what stood out when I read the sources. Five of the eight list keynote speaking on their own page, and most of the rest publish a newsletter. Daily short video, the format that dominates American lists, barely features. So AI influence on Australian business mostly travels through conference stages and inboxes, and that changes how you use these people. You don't scroll past them. You subscribe, or you book them.

It changes the timing too. A newsletter turns up on its own schedule and asks for ten minutes of your attention. A keynote is one afternoon a whole leadership team sits through together, which is why it's often the better way to shift a room's thinking. If your problem is that the board hasn't taken AI seriously, a speaker beats a feed. If your problem is that you can't keep up yourself, an inbox habit beats a speaker.

Five kinds of voice and the job each one does

The researcher. A professor or an institute's chief scientist works on a timescale of years. Follow one for perspective: what's hype, what's real, and what the next few years might bring. Don't expect tool picks. The value is that their claims have usually been through peer review before they reach you.

The ethicist. Someone who advises governments and boards on responsible technology asks the questions your legal and HR people will ask later, only earlier. That's worth a lot to a leadership team about to put AI near customers or staff data, and very little to a sole trader choosing a writing tool.

The founder-operator. A person running a business on AI, or running an AI company, gives you a worked example. It's the most transferable kind of advice and the easiest to misuse, because their sector, size and appetite for risk may be nothing like yours. Copy the method, not the conclusion.

The futurist. Executive framing for leadership retreats and conferences: what AI means for the business model, not which button to press. Good for a strategy day, thin for Monday morning.

The tracker and the tool tester. One catalogues models and datasets so you know what exists; the other tries tools and sorts them by the job they do. Both save you hours. Tool testers often have brand partnerships, and the good ones say so openly. Read their picks as editorial with commercial ties, and run your own trial before you pay for anything.

Check the Australian tie yourself

Only four of the eight state an Australian tie on a page I could read, and I haven't given anyone else an Australian address. For following someone, location barely matters: a sharp newsletter reads the same in Perth or Porto. For booking, it does. Travel, time zones and whether a speaker knows the local regulatory picture and the kind of businesses in the room all come into it. When you shortlist a speaker, look for the tie on their own page, not on a round-up, this one included.

An hour-long audit for any AI voice

Before you add anyone to your reading list or your events budget, give them an hour. It's the cheapest filter there is.

1. Open their own site, not a profile on someone's list. Write one line each on what they do, who it's for and how they're paid. If you can't fill all three, that's your first answer.
2. Read their last three newsletters, articles or talk descriptions. Count how many gave you something you could try this week.
3. Look for disclosure. Anyone recommending tools should say where they have commercial ties.
4. Take one idea and spend twenty minutes trying it on a real task. Write down what happened.
5. Score them: three for something you'd keep doing, two for something useful to know, one for noise. Follow the threes for a month and let the rest go.

Do it for one researcher, one operator and one practical source, and you've got a balanced feed for a month's trial instead of a list of famous names.

Following costs time; booking costs money

Following any of these people costs attention, and attention is the real budget: three newsletters you never open are worse than one you read. Booking is different. Speaker and workshop fees in this field are usually quoted on request, so there's no figure I can stand behind here. What drives the fee is clear enough: the length of the session, travel, how far the talk is shaped around your organisation, and whether there's a workshop or Q and A on top. Ask for the fee and what's included in one email, and ask whether they'll work with your team's own examples or deliver a standard talk.

Traps when you build an AI reading list

The first is following too many. Ten voices means skimming all of them; three means acting on one. The second is mistaking a big stage for relevance. A brilliant keynote on the future of work may have nothing to say about your accounts team's Tuesday.

The third is treating a founder's story as a template. A lean business run on AI agents is a useful proof that it can be done, not a plan for a firm with forty staff and a compliance team. The fourth is assuming a researcher will train your people. Researchers explain; trainers change habits. And the fifth is ignoring disclosure: a tool tester who never mentions partnerships isn't more neutral, just less open.

Eight Australian AI voices worth reading

Toby Walsh

A UNSW Sydney professor of artificial intelligence and chief scientist of its AI institute, for a researcher's read on where the field is going.

Catriona Wallace

A responsible-AI ethicist and book co-author, suited to boards that want the governance questions raised before anything reaches customers.

Marek Kowalkiewicz

A Queensland University of Technology chair in digital economy, working where AI meets government and industry transformation.

Alan D. Thompson

A model-tracking data voice whose LifeArchitect.ai tables and The Memo catalogue which AI models and datasets exist.

Nici Sweaney

A scientist turned founder of Ai Her Way, teaching the AI operating system she runs her own business on, with women leading in AI as the aim.

Amanda Johnstone

Chief executive of AI and mental health technology company Transhuman, speaking and advising leadership teams on people-first AI.

Mark van Rijmenam

A futurist who speaks at leadership retreats and conferences and writes the Synthetic Minds newsletter on what AI means for business.

Sam Despo

A Melbourne agency owner and creator behind a searchable AI tools directory and newsletter, organised by the job you need done.

A British practitioner on an Australian list

Because I picked the names, mine included, treat this entry as the least neutral on the page. I'm not Australian and I don't claim an Australian base. I'm British, and I work in English with clients worldwide, in person or live online on UK hours. I'm not a researcher or a professor, I don't train models, and I don't publish a daily news feed or a video channel. The people above do those things, and you should follow them for it.

What I bring is the practitioner's view. I've worked in marketing for 23 years and went all in on AI in 2024, rebuilding my own consultancy around it. I've spoken at more than 100 events, including TEDxBolton and Oxford University. My sessions are built on a team's own tasks and end with a one-page agreement on safe AI use, whether that's a 90-minute workshop, a keynote of 45 to 60 minutes or a short series over a few weeks. I've written about AI training for Australian teams separately, and there's a page for event organisers booking an AI keynote in Australia too.

I'd rather show than tell. Between January and September 2026, my own site went from 1,322 Google clicks a month to 6,412, and visits from AI assistants went from 32 a month to 3,599, measured in my Search Console and GA4. My newsletter reaches an audience of 15,000.

My pick by kind of voice

One researcher or ethicist for the big questions, one founder or operator for a lived example, and one practical source, either a tool tester or a model tracker. After a month, keep the one that changed something you did and drop the rest.

If you'd rather have an expert than a voice to follow, my page on finding an AI expert in Australia covers that.

Reading the right voices sharpens your thinking, but it won't move work off anyone's desk on its own. I run AI training and workshops for teams, and help businesses with AI enablement and workflow automation, so the busywork runs itself and people get time back. Most of that work is with SaaS marketing teams and small businesses. Here's how we can work together.