Imagine walking into a restaurant and the menu on the screen knows your dietary preferences, remembers what you ordered last time, and suggests dishes based on your past behaviour. That’s not a hypothetical — it’s already being tested in several Australian restaurant groups.

What’s happening

A handful of multi-venue restaurant chains in Australia are piloting AI-powered menu personalisation systems. The technology works through loyalty apps or QR code ordering systems that identify individual customers and adjust what’s presented to them.

The simplest version filters out items containing allergens the customer has flagged. More advanced versions highlight dishes similar to previous orders, suggest new items based on taste preferences inferred from order history, and even adjust pricing for loyalty rewards.

One Melbourne restaurant group I spoke with has been running a pilot for four months. Their system analyses order history across all venues and creates a personalised “recommended” section at the top of the digital menu for returning customers. They’ve seen a 15 percent increase in average order value from customers who engage with the recommendations.

The technology

The underlying systems combine recommendation engines (similar to what Netflix and Spotify use) with restaurant-specific data: order history, time of day patterns, seasonal preferences, and spending behaviour.

Some systems also incorporate real-time factors like current kitchen capacity and ingredient availability. If the kitchen is backed up on grills but the pasta station is quiet, the AI might promote pasta dishes more prominently to new customers. If an ingredient is running low, items using it get deprioritised.

The data integration required is significant. These systems need to connect POS data, inventory management, kitchen display systems, and customer databases. AI agency Sydney firms and similar consultancies are building these integrations for hospitality clients, connecting multiple data sources into unified recommendation engines.

What’s good about it

For customers with dietary restrictions, personalised menus are genuinely helpful. If you’re coeliac, having the menu automatically filter to gluten-free options saves the awkward conversation with every new waiter at every new venue.

For restaurants, the efficiency gains are real. Better demand prediction means less waste. Higher engagement with menu items means better average spend. And the data feedback loop helps chefs understand what their customers actually want versus what they assume they want.

The allergy management angle is particularly important. Allergen-related incidents in Australian restaurants remain stubbornly high despite training programs and regulations. A system that automatically flags and communicates allergen information based on customer profiles adds a genuine safety layer.

What concerns me

The privacy implications are obvious. A system that knows what you eat, when you eat it, how much you spend, and what your dietary preferences are holds intimate personal data. How that data is stored, shared, and potentially sold matters.

Australian privacy law requires businesses to be transparent about data collection and use. But the practical experience of digital services in other industries suggests that transparency is often buried in terms and conditions nobody reads.

There’s also the filter bubble problem. If the AI keeps showing you dishes similar to what you’ve ordered before, you might never try something unexpected. Part of the joy of dining out is discovering dishes you didn’t know you’d love. A system optimised for conversion (showing you what you’re most likely to order) works against culinary exploration.

And then there’s the pricing question. Dynamic pricing based on customer data is technically possible with these systems. Showing different prices to different customers based on their spending history is a short step from personalised recommendations. I asked the Melbourne restaurant group about this directly. They said they weren’t doing it and had no plans to. But the capability exists.

What about the staff

Personalised digital menus reduce the role of waitstaff as guides to the menu. A good waiter who knows the menu, reads the table, and makes genuine recommendations based on conversation is worth more than any algorithm. The risk is that AI personalisation becomes a reason to further deskill front-of-house roles, turning waiters into order carriers.

Some restaurant operators see it differently. They argue that freeing waitstaff from routine menu explanations allows them to focus on the hospitality aspects — welcoming guests, managing pace, creating atmosphere — that machines genuinely can’t do.

Where this goes

Menu personalisation is coming. The technology works, the economics make sense for restaurants, and customers (particularly younger ones) expect personalised experiences from every digital interaction.

The question isn’t whether it will happen but how it’s implemented. Done well, with strong privacy protections, transparent data use, and a genuine focus on customer benefit, it could make dining better. Done poorly, it becomes another way to extract maximum spend from customers while eroding the human elements that make restaurants worth visiting.

I’ll be watching this closely. And I’ll keep asking for the paper menu.