The backup system sabotaged by the network it was meant to replace
For the second day running, Australians reaching for the phone in moments of crisis found the line dead — not from malice, but from a single misfiring clock deep inside Telstra's infrastructure. A GPS node that lost its sense of time unravelled a chain of dependencies stretching from emergency dispatch to regional railways to corner-shop payment terminals, reminding us how thoroughly modern life is threaded through systems whose fragility we rarely see until they fail. The episode joins a growing ledger of outage-linked harm in Australia, and now sits before regulators asking an old question in a new form: when the infrastructure of safety is privately held, who answers when it breaks?
- Hundreds of Australians dialling triple zero on Thursday morning heard only error tones — the outage Telstra had declared resolved was still quietly unravelling.
- A single GPS node reset had corrupted time synchronisation across the entire network, cascading into failed calls, halted trains, and silent Eftpos terminals for small businesses.
- Victoria's regional rail network seized during peak hour as Telstra's 4G signal ironically jammed the backup satellite phones trains depend on when mobile coverage fails.
- Telstra completed 639 welfare checks, but seven people needed emergency referrals and 170 cases were handed to police — the human cost behind the corporate apology.
- Government launched a formal investigation with penalty powers, and new ACMA rules now compel telcos to publish outage timelines publicly, making Telstra's failures part of the permanent record.
For the second consecutive day, Australians trying to reach triple zero found themselves locked out. On Thursday morning, hundreds of Telstra customers dialled the emergency number only to receive an error message. The company had promised the problem was fixed. It wasn't.
The root cause traced back to a software fault in Telstra's GPS node — a server responsible for keeping the entire network synchronised to the nanosecond. When it reset, it corrupted time data across the whole system. The domino effect was swift: calls failed to route, Victoria's regional trains ground to a halt during peak hour, and businesses couldn't process payments. The cruel irony of the rail disruption was that Telstra's 4G signal was interfering with the backup satellite phones trains rely on when mobile coverage fails — the safety net sabotaged by the very network it was meant to replace.
Telstra's chief financial officer Michael Ackland apologised publicly, then urged customers whose first emergency call failed to try again immediately — a workaround that amounted to a quiet admission the system remained unreliable. The company conducted 639 welfare checks; seven people required emergency assistance, and 170 cases were referred to police for further assessment.
The government moved to signal accountability. Communications Minister Anika Wells noted she had heard of no deaths or serious injuries, a careful statement that acknowledged the stakes without closing the door on concern. Industry Minister Tim Ayres announced a full investigation with the threat of financial penalties. Opposition leader Angus Taylor, who had initially floated a link to a Chinese missile test in the Pacific, defended the speculation as reasonable given the information vacuum at the time. Ackland was unequivocal: this was not a cyber attack.
The failures landed against a difficult backdrop. An Optus outage in September 2025 had been linked to two deaths. Consumer advocates called for compensation, and new rules from the Australian Communications and Media Authority now require telcos to publish detailed outage timelines publicly. Telstra's cascade would become part of that record — examined, dissected, and measured against the question of whether transparency alone is sufficient to prevent the next collapse.
For the second consecutive day, Australians trying to reach triple zero found themselves locked out. On Thursday morning, hundreds of Telstra customers dialed the emergency number only to receive an error message as their phones searched for an alternative network that wasn't there. The company had promised the problem was fixed. It wasn't.
This was the aftermath of Wednesday's catastrophic network failure—a software glitch in what Telstra calls its GPS node, a server that tells the entire network what time it is, down to the nanosecond. When that server reset, it corrupted the time synchronization across the whole system. The domino effect was immediate and brutal: calls couldn't route properly, trains couldn't run, businesses couldn't process payments. By Thursday, the same underlying defect was still causing havoc, just in a different way.
Telstra's chief financial officer Michael Ackland stood before the public and apologized, then immediately tried to reassure them. "Customers can feel confident in calling triple zero," he said, even as the evidence suggested otherwise. The company urged people whose first call didn't connect to try again immediately—a tacit admission that the system was unreliable, dressed up as a workaround. The telco had completed 639 welfare checks since the outages began. Of those, seven people actually needed emergency assistance, their details passed to emergency services. Another 170 cases went to police for further assessment. Two hundred thirty people responded to welfare checks by text, meaning they were well enough to do so. The math of crisis management.
The chaos extended far beyond the phone lines. Victoria's entire regional rail network ground to a halt during Thursday morning's peak commute. Telstra's 4G signal was interfering with the backup satellite phones that trains rely on when the mobile network fails—a cruel irony, the backup system sabotaged by the very network it was meant to replace. Trains in New South Wales were also disrupted, though services gradually resumed on the Southern Highlands and Hunter lines. On Wednesday, small businesses across the country had been unable to take customer payments because their Eftpos terminals couldn't connect. The economic bleeding continued even as the network came back online.
The government moved quickly to signal accountability. Communications Minister Anika Wells said she had heard of no deaths or serious injuries linked to the triple zero failures, a careful statement that acknowledged the stakes. Industry Minister Tim Ayres announced a full investigation with the threat of penalties. But the political response was muddied. Opposition leader Angus Taylor, who had initially suggested the outage might be linked to a Chinese missile test in the Pacific, defended that speculation as reasonable given the lack of official explanation. Ackland was emphatic on one point: this was not a cyber attack.
The broader context made the failures harder to swallow. In September 2025, an Optus outage lasting nearly 14 hours had been linked to two deaths. In June, Vodafone customers experienced widespread reception and data problems. The Australian Communications Consumer Action Network's chief executive, Carol Bennett, called for compensation. "There certainly needs to be some kind of recourse," she said, "to compensate customers and businesses who've lost a lot when it comes to these outages." In March, new rules from the Australian Communications and Media Authority had required telcos to publish detailed timelines of when outages started and stopped, along with their causes. Telstra's failures would now be part of the public record, examined and dissected. The question now was whether transparency alone would be enough to prevent the next cascade.
Citas Notables
Customers can feel confident in calling triple zero. No system is completely immune from impacts and outage.— Michael Ackland, Telstra CFO
There certainly needs to be some kind of recourse to compensate customers and businesses who've lost a lot when it comes to these outages.— Carol Bennett, Australian Communications Consumer Action Network
La Conversación del Hearth Otra perspectiva de la historia
Why does a GPS server resetting cause the entire network to fail? That seems like a single point of failure.
Exactly. The GPS node tells every other system in the network what time it is—to the nanosecond. When it reset, it sent the wrong time everywhere at once. Everything downstream depends on that synchronization. It's like if the clock in a train station suddenly jumped forward six hours; every schedule breaks.
And they couldn't just fix it on Wednesday?
They did fix it. But the underlying software defect was still there. Thursday's problem was the same fault manifesting differently. They had to address it in a completely different way the second time.
What strikes me most is the welfare checks—639 of them, with 7 needing actual emergency help. That's a lot of people in distress.
Yes. And 170 cases went to police. We don't know what those were—elderly people who couldn't reach help, someone having a medical emergency, someone in danger. The company did the checks after the fact, which is good, but it doesn't undo the hours people spent unable to reach emergency services.
The backup system for trains was sabotaged by the main network. How does that happen?
The satellite phones on trains are supposed to work when mobile networks fail. But Telstra's 4G signal was so disrupted it was interfering with those backup systems. So the redundancy failed. There was no fallback.
What does Ackland's insistence that people can "feel confident" in triple zero actually mean?
It means the system itself is robust—the infrastructure, the routing, the protocols. But confidence requires reliability, and they just proved it's not reliable when their network fails. He's trying to separate the emergency system from the network that serves it. The problem is, they're not separate.