Why Australian Quick-Service Restaurants Are Putting AI Behind the Menu Board in 2026
KFC Australia is trialling voice AI at the drive-thru in five NSW restaurants. Hungry Jack’s has a voice-activated ordering system running at St Peters in Sydney.
The screens are next. Software that decides what your digital menu boards show, and when they show it, is already running in trials at the biggest QSR groups in the world. What it does is narrower than the name suggests, and it only works when the systems underneath it already do.
What does AI do behind a digital menu board?
It picks what the board shows, using live inputs like time of day, stock levels, sales patterns and weather.
The AI chooses between options your team has already built, which caps its usefulness at two things: the quality of the data feeding it, and the range of templates it can choose between. Give it three templates and stale stock data, and it will make a worse decision than your shift manager.
The largest global QSR groups have been running AI menu board trials for a couple of years. In the trials reported publicly so far, the systems adjusted layout and item order while leaving the pricing structure alone, which puts the working function at ranking and emphasis.
These are small trials, often a handful of restaurants each, run by operators with more resources than anyone else in the sector. The technology is early even at the top end.
Most of what gets labelled AI on a menu board is closer to sophisticated dayparting. Dayparting means showing different menus at different times, so breakfast gives way to lunch without anyone touching a screen. The AI version adds inputs beyond the clock, so a cold Tuesday morning can push soup up the board while a heatwave promotes iced coffee instead.
Why are Australian QSRs looking at this now?
Digital ordering already dominates at the major chains, so operators finally hold the data that automation needs.
More than half of orders across Australia’s major fast-food chains now come from digital sources, with app ordering a significant share on top of counter and kiosk traffic. That volume produces something Australian operators did not have five years ago, which is a per-store, per-hour record of what sells, to whom, and when.
Menu automation is only as good as the demand data behind it, so that shift is what moved AI menu boards from a conference demo to a budget conversation. Australian interest in them follows the order data, and the order data only arrived recently.
The market it is landing in is growing. GapMaps, a location-analytics firm that publishes an annual report on the sector, recorded 7,779 locations across Australia’s top 36 QSR brands in the 12 months to December 2025, with 359 openings against 109 closures for a net gain of 250 stores. That is the strongest year of net growth in a decade, and the largest brands drove most of it, with several opening 25 or more stores each.
More stores means more screens, more menu variations and more manual updates, which is the workload AI is being pointed at. If you run QSR digital signage across a growing network, the administrative load grows with it.
What can AI change on the board, and what should it?
Reordering items, swapping promotions and hiding sold-out stock are safe wins. Pricing needs more care.
Dynamic pricing gets most of the attention, while the value most Australian operators can bank sits with availability and emphasis. A stock-aware board removes manual work with no downside for the customer, since all they experience is accurate information sooner. When an item runs out, it greys out or disappears without anyone walking to a screen.
Price changes carry more weight, because they touch customer trust and disclosure obligations at the same time. Hungry Jack’s AI drive-through trial drew critical public coverage focused on job displacement, so Australian customers demonstrably notice AI at the counter and form a view about it. Whatever you automate at the price line, your reputation is part of the sum.
Compliance sets a limit on all of it. If you operate 20 or more outlets in NSW, Victoria, Queensland, South Australia or the ACT, or 50 or more nationally, your menu boards must display kilojoule values, and the rules cover drive-through boards and digital frames explicitly. Anything the AI changes has to leave that information intact and correctly sized, so the kilojoule value must travel with the item through every re-rank, which is a template question before it is an AI one.
Check the NSW Food Authority requirements and your own state’s rules before you automate anything that moves an item or a price.
The safest starting list looks like this:
- Stock-aware hiding. Item sells out, item comes off the board.
- Daypart transitions. Breakfast to lunch to dinner, without staff intervention.
- Promotion rotation. Feature slots that change by time, day or conditions.
- Weather-triggered emphasis. Hot drinks up on cold mornings, cold drinks up on hot ones.
Each of those is measurable, reversible and invisible to the customer as anything other than a menu that happens to be right. None of them requires you to explain a pricing algorithm to a queue.
What has to be working before AI is worth switching on?
Clean POS data, a disciplined template system, and a content management system that can act on both.
Most of the work sits here, before any algorithm runs. AI picks between options because someone built those options as templates and fed it live data, so without template architecture and point-of-sale integration an AI board has nothing to decide between and no basis on which to decide.
Template discipline means your menu is built as a system. Items are data, prices are data, kilojoule values are data, and the layout is a rule that arranges them, which is what lets a machine rearrange the board safely. Most networks are still running flat images, where nothing moves without a designer.
POS integration is the other half. Your point-of-sale system already knows what sold, what ran out and what it costs, so connecting it to your screens is what turns a display into something that can respond. This is the foundation work, and it takes longer than the AI layer that sits on top of it.
Scale makes both harder and more valuable. Amped Digital manages 4,795+ screens for Subway across Australia and New Zealand, and upgraded the digital menu signage across 30+ Krispy Kreme locations in ANZ. Networks of that size only stay consistent because the templates and the data pipeline are built to hold, which is the same groundwork any AI layer would depend on.
If your foundation is not there yet, that is the work. Our content design and project management teams spend most of their time on exactly this, and it pays off whether or not you ever switch on an algorithm.
How should a multi-site operator start?
One automation, one daypart, one group of stores. Prove it, then widen.
The temptation with a network is to solve it once and roll it everywhere, and that is the approach most likely to produce an expensive reversal. A single rule on one daypart in a handful of stores stays measurable and reversible, because the rest of your network is sitting right there as a control group. You find out whether the automation helped before you commit the estate to it.
Pick the automation with the clearest before-and-after. Stock-aware hiding is a good first candidate, since the outcome is easy to see and nobody has to be persuaded that an accurate menu is better than an inaccurate one.
Give it long enough to survive a full trading cycle, including a weekend and a quiet midweek. Then compare against your untouched stores on the numbers you already track: order accuracy, speed of service, average order value, staff hours spent on menu admin.
The same logic applies as you extend outward to drive-thru digital menu boards and self-service kiosks, where the stakes are higher because a mistake sits in front of a queue of cars. If you are working out how to keep content consistent while you do this, our guide to managing digital menu boards across multiple locations covers the governance side.
The practical view for 2026
AI behind digital menu boards works, within limits worth knowing before you budget for it. Three things are worth carrying out of this.
The technology only selects from the templates and data you give it, so the quality of your inputs sets the ceiling on the output. The automations that pay are the unglamorous ones, because availability and emphasis deliver returns without asking your customers to trust an algorithm with the price. Underneath both sits the foundation work of clean POS integration and a proper template system, which earns its keep whether or not you ever switch an algorithm on.
That ordering is the whole point. Get the plumbing right this year and you can adopt whatever the technology grows into, because the groundwork does not change when the algorithm does.
If you want a straight assessment of whether your menu board network is ready for any of this, let’s talk.