How facial recognition makes self-exclusion actually work

Self-exclusion is one of the most important tools in responsible gambling. It lets a person who recognises they're at risk make a decision for their future self: bar me from this venue, even when I'm tempted to come back. But a self-exclusion is only as good as a venue's ability to honour it, which is exactly where it has traditionally fallen down.

The problem with relying on memory

For decades, enforcing self-exclusion has meant asking staff to remember faces from a folder of photographs and spot them in a crowd. On a quiet afternoon, maybe. On a busy Friday night, across hundreds of patrons and shift changes, it's close to impossible. People who've excluded themselves slip back in, not because anyone failed, but because human recognition simply doesn't scale. The result is a program that looks good on paper and leaks in practice.

Where facial recognition fits

Facial recognition closes that gap by doing the first check automatically. As a patron enters a venue or a gaming area, the system compares their face against the self-exclusion list and, if there's a potential match, alerts staff in real time. It doesn't get tired or look away, so the first check holds up at peak just as it does at open.

A match isn't an automatic ejection. It is a quiet prompt for a trained person to confirm identity then handle the situation with care and discretion.

Crucially, the technology supports staff rather than replacing them. The NSW Code of Practice is explicit on this: facial recognition supplements existing processes; the human stays in charge of the decision.

Why timing matters

The earlier a match surfaces, the better the outcome. Detecting a self-excluded patron at the entrance, before they've sat down at a machine, means the interaction can be calm, private and supportive, rather than a confrontation mid-session. Speed isn't about catching people out; it's about intervening gently and early, which is better for the patron and for staff alike.

Doing it without creating a surveillance problem

Understandably, people worry about facial recognition scanning everyone who walks in. A well-designed system answers that concern in its architecture. At Ottica, detection runs at the venue and data stays in Australia. If you're not on the exclusion list, your biometric data is discarded on the spot. The only biometric information we keep belongs to people on the list. The point is narrow and specific: support self-exclusion, nothing more.

Accuracy you can stand behind

For a self-exclusion program to be trusted, the matching has to be reliable in two directions at once: catching the people who asked to be caught, without wrongly flagging someone who didn't. Ottica is built on a face-matching engine that performs strongly on independent industry benchmarking, paired with liveness detection so a printed photo or a phone screen won't pass as a real person.

• Thresholds are tuned for venue conditions rather than left at a lab default.
• Performance is monitored against the false positives and false negatives the Code asks you to record.
• Every alert goes to a trained person rather than triggering an automatic response.

That combination is what keeps false positives rare and the patron experience respectful.

The bigger picture

Used well, facial recognition turns self-exclusion from a good intention into a working safeguard. It gives people who've made a hard decision a far better chance of sticking to it, plus it gives venues a credible way to meet their responsible-gambling obligations. The NSW Government has committed to mandating facial recognition in hotels and clubs with gaming machines once the statewide exclusion register is operating, so that capability is moving from nice-to-have to expected.

If you'd like to see how this works on a real gaming floor, we're happy to show you.

Current as at July 2026. This article is general information, not legal advice.

Sources

Liquor & Gaming NSW — Code of Practice: Facial Recognition Technology in Hotels & Clubs. Lander & Rogers — NSW takes a governance-first approach to facial recognition technology.