AusAlert Test Jolts Cinemas but Leaves Phones Silent

Australians were told almost every compatible device would scream at 2pm AEST, but the first nationwide AusAlert test landed unevenly: loud in parks, libraries and cinemas, silent for some phones sitting switched on.

The federal government called the drill a success after 94% of targeted cell towers broadcast the alert, according to Guardian World. The test message was meant to show “TEST MESSAGE” and “NO ACTION IS REQUIRED”, alongside a siren-like warning on compatible phones, some smartwatches and tablets.

AusAlert test sounded across Australia at 2pm, but not for everyone

The AusAlert test was the first national run of Australia’s new emergency warning system, which is scheduled for deployment in October for fires, floods, terrorist incidents and health emergencies.

The signal was supposed to reach compatible devices across the country at about 2pm AEST. For many people, it did exactly that. It blared through shared spaces, interrupted routines and lasted only seconds.

For others, the moment barely registered.

“I was counting down for two minutes before,” Canberra worker Indigo Miller told Guardian Australia. “But it was pretty anti-climactic and I think I accidentally clicked it away.”

Miller said her alert was not especially loud. She added: “Mine also wasn’t super loud so if I was outside and my phone was inside, I don’t know if I would hear.”

The experience was sharper in public places. Busker Dani Bainbridge was preparing to perform at Sydney’s Hyde Park when devices began pinging.

“It did the job, I think,” she said.

At the State Library of NSW, the siren cut through the usual silence, giving Year 12 students a break from study and waking one high school student who had been nodding off at a laptop. In a Sydney Imax screening of Christopher Nolan’s The Odyssey, some phones still sounded despite workers telling ticket holders to switch them off before the 1.50pm session. The alerts rang during ads before the film.

For readers tracking how public spaces are being pulled into technology flashpoints, XOOMAR has also covered Avoiding AI Workshops Turn Libraries Into Big Tech Revolt and Big Tech Blocks Digital Services Act Data Access in EU Test. The AusAlert drill was different in purpose from the Enigma robot control public test, but similar in one respect: the real test happened in messy public settings, not in a lab.

Government calls AusAlert test a success despite silent phones

Emergency Management Minister Kristy McBain said 94% of targeted mobile towers successfully received and broadcast the signal. That is the government’s strongest number so far.

It is not the same as saying 94% of people received the warning. The government has not released a device-level delivery rate in the supplied material.

McBain urged people not to assume a missed alert meant the whole system had failed.

“Don’t worry if you did not receive an alert – this could be because you were not in a coverage area, your software was not up to date or your device was incompatible,” she said.

She also said the National Emergency Management Agency would keep reviewing technical and performance data and would work with emergency service organisations across the country.

Asked whether AusAlert could ever be completely successful, McBain was direct.

“There is no one avenue of communication that is 100% successful.”

She said the new system is “more reliable than the current text-based technology”, while framing it as one more channel for emergency agencies rather than a replacement for every other warning method.

The compatibility rules matter. Before the test, NEMA had advice that about 80% of Apple iPhones would ring, along with 16 million Android devices.

Guardian Australia reported accounts of an iPhone 15 Pro, iPhone 17 Pro, iPhone 12, a Samsung phone and other devices not receiving a notification despite being on. It also received reports from users whose phones displayed a notification but did not make the loud sound.

That gap is the tension. AusAlert is designed for moments when seconds matter. A national test can tolerate unevenness if engineers learn from it. A bushfire, flood or fast-moving public safety incident gives authorities less room.

Critical alerts exposed a safety problem for hidden phones

The test also raised a risk outside the usual engineering checklist.

Domestic and family violence advocates warned before the drill that the alert could expose “safe phones”, meaning hidden devices used by people trying to stay connected without being monitored. AusAlert interim coordinator general Pat Hetherington said the advice for people with safe phones and hidden phones was to switch them off for a 24-hour period.

That advice shows how a technical rollout can create a personal safety issue even when the alert works as designed.

The government’s own AusAlert material says the system uses cell-broadcast technology over active 4G and 5G networks and is designed to work regardless of carrier. It also says compatible devices do not need a working SIM card to receive alerts, if they can connect to a network.

The same material says assistive technology results have varied. Some configurations involving cochlear implants, hearing aids and refreshable braille displays worked well, while others behaved inconsistently.

Before the test:

• Expectation: compatible devices would receive a loud critical alert at about 2pm AEST.
• Public preparation: the government warned Australians in advance.
• Safety advice: people with hidden phones were told to switch them off for 24 hours.

After the test:

• Government metric: 94% of targeted cell towers broadcast the signal.
• Public reports: some devices made no sound, some got no notification, some alerts lasted only seconds.
• Open question: no released count yet shows how many people missed the AusAlert test.

October rollout now depends on explaining the misses

The next task for officials is not just proving that towers broadcast the alert. It is explaining who did not receive it, and why.

Based on the supplied statements, the confirmed possible reasons include coverage gaps, outdated software and incompatible devices. The government has also pointed to continuing analysis by NEMA, which should matter more than the headline success figure if AusAlert is going live in October.

XOOMAR analysis: the drill gave emergency agencies something useful because it produced failures in the open. That is uncomfortable, but better than discovering the same issues during a real evacuation or health emergency.

The practical questions now are narrow and answerable: how many compatible devices missed the alert, whether the failures clustered around particular device models or software versions, and what users should do before October beyond installing updates and restarting devices.

The AusAlert test lasted seconds. The follow-up should not. Australia now has a national warning system with a public benchmark, a visible failure set and a deadline. The October launch will be judged on whether officials turn those missed pings into clear fixes before the next emergency forces the system out of rehearsal.

Impact Analysis

• The uneven test shows emergency alerts may not reliably reach every compatible device in a real crisis.
• AusAlert is scheduled for deployment in October for fires, floods, terrorist incidents and health emergencies.
• Public testing helps identify gaps before the system is used for life-threatening events.