We knew this could happen and didn't prevent it
On a Wednesday morning in Australia, Telstra's nationwide network collapsed for nearly five hours after a software fault reset the company's internal clock nearly two decades into the past — a known vulnerability that had felled another carrier before, and one the government had already flagged as critical. CEO Vicki Brady returned from overseas to offer an apology, but the harder question lingering beneath the technical postmortem is whether foreknowledge of a risk, without adequate remedy, carries a moral weight that no apology can fully discharge. In the shadow of that question, South Australian police are investigating whether a failed emergency call during the outage may have contributed to a death — a possibility Telstra disputes, leaving the truth to be weighed by a coroner.
- A single technician's work on a time-keeping node sent Telstra's network clock spiraling back to November 2006, collapsing mobile, payments, rail, and emergency services across an entire nation in minutes.
- The outage exposed a painful irony: Telstra had been warned by the government's own infrastructure security body about the fragility of its timing systems, yet the redundancies built to prevent exactly this failure did not activate.
- The same clock-reset vulnerability had already destroyed half a Jersey telecoms network in 2020 — a documented precedent that makes Telstra's unpreparedness harder to explain away as mere bad luck.
- South Australian police are now investigating whether a person's death is linked to a triple-zero call that allegedly failed to connect during the outage, a claim Telstra flatly contests with its own technical records.
- With a coroner set to weigh competing accounts — police testimony, family experience, and Telstra's network logs — the outage has moved from corporate crisis into potential legal and ethical reckoning.
Vicki Brady returned from an interrupted overseas holiday to face the public and account for how Telstra's network — carrying mobile calls, payments, rail systems, and emergency services for an entire country — had gone dark for nearly five hours. The cause was precise and almost absurd: a software fault in the company's time-keeping systems had reset the network's internal clock to November 2006, setting off what experts described as a cascading digital collapse that unravelled the infrastructure within minutes.
Brady apologised, but the apology was complicated by what had already become known. The government's Cyber and Infrastructure Security Centre had previously warned Telstra about the critical role its timing systems played in network resilience. Brady herself acknowledged that timing infrastructure was well-understood to be essential. Yet when the fault struck, the redundancies designed to absorb exactly this kind of failure did not hold. "Our network is built with a lot of redundancy," she said, "and that will be part of this investigation" — an admission that the safety net had simply not worked.
The failure was not without precedent. In 2020, a telecommunications company in Jersey lost roughly half its network when a time server generated an incorrect date — November 2000 — which propagated through routers that then isolated themselves. Recovery took nearly five days. The root cause was a week counter that maxed out and defaulted nearly two decades into the past. Telstra's outage followed the same pattern, the same flaw, the same known risk.
Beyond the disruption to commerce and transport, the outage carried a graver shadow. South Australian police opened an investigation into whether a death during the blackout was connected to the network failure. A person had been taken to a regional hospital, and their spouse had attempted to call triple zero for an ambulance — a call that, according to police commissioner Grant Stevens, did not connect, forcing the family to find another phone. A coroner's report would follow.
Telstra disputed the account directly. Chief financial officer Michael Ackland said the company found no record of calls from the relevant numbers reaching its network to access triple zero, that no outage was active in the local area at the time, and that mobile signal at the location had been strong. A separate triple-zero call from a nearby number had gone through without issue. The contradiction remained unresolved — police pursuing a potential link between the outage and a death, Telstra insisting its records showed no failure — with the coroner left to determine what the evidence, and the human cost, ultimately meant.
Vicki Brady stood before the public for the first time since cutting short an overseas holiday, and the weight of the previous Wednesday morning sat visibly on her shoulders. Telstra's chief executive had come to explain how her company's network—the one that handles mobile calls, payments, rail systems, and emergency services across an entire nation—had simply stopped working for nearly five hours. The cause was almost mundane in its specificity: a software fault in the company's time-keeping systems had reset the network's internal clock to November 2006, triggering what one expert described as a "digital domino chain fall" that collapsed the infrastructure in minutes.
Brady's apology was direct. She understood, she said, the broader impact when services went down—the payments that couldn't process, the trains that stopped, the disruption that rippled through the community. She was sorry. But what emerged in the days following the outage was something more troubling than a simple technical failure: Telstra had known the risks. The government's Cyber and Infrastructure Security Centre had previously warned the company about the critical nature of its time-keeping services to infrastructure resilience. Brady acknowledged that timing systems were "very well-known" and "critical" in mobile networks. Yet when a technician worked on a time-keeping node and it reset, the backup systems that should have prevented catastrophe simply did not. "Our network is built with a lot of redundancy put in place, and that will be part of this investigation," Brady said, unable to explain why those redundancies had failed.
Telstra was not the first to stumble into this particular trap. A telecommunications company in Jersey had suffered an identical outage in 2020, when its time server generated the wrong date—November 2000, nearly twenty years earlier. That incorrect date propagated through the network's routers, which then isolated themselves, causing the company to lose roughly half its network and ultimately triggering a complete outage that took nearly five days to fully recover. The investigation into that failure revealed the underlying vulnerability: the system used a week counter that maxed out at 1,023 weeks, and the default date reset to 19.6 years in the past. Telstra's failure followed the same pattern, the same weakness, the same known risk.
But the outage's consequences extended beyond service disruption and corporate embarrassment. South Australian police opened an investigation into whether a death during the outage could be linked to the network failure. A person had been taken to a regional hospital, and the person's spouse had attempted to call triple zero for an ambulance. That call did not go through, according to police commissioner Grant Stevens, who said the family had to use another phone instead. The police would prepare a report for the coroner, Stevens said, and the coroner would ultimately determine what implications the failed call carried.
Telstra disputed the account. Michael Ackland, the company's chief financial officer, said on Friday that Telstra had found no record of calls from the relevant phone numbers accessing Telstra's mobile network to reach triple zero. The company also confirmed there were no active outages affecting the local area at the time and that mobile signal strength at the location had been good. A related call to triple zero from another number had gone through successfully. The contradiction hung unresolved: police investigating a potential connection between the outage and a death; Telstra insisting its records showed no failed call. The coroner would have to sort through the competing claims, the technical data, and the human cost of a network that had been built with redundancy but had failed when it mattered most.
Citas Notables
Our network is built with a lot of redundancy put in place, and that will be part of this investigation— Vicki Brady, Telstra CEO
To date, we can see no record of calls from those numbers accessing Telstra's mobile network to call triple zero— Michael Ackland, Telstra CFO
La Conversación del Hearth Otra perspectiva de la historia
Why does a time-keeping system matter so much to a mobile network? It seems like a small thing.
It's not small at all. Every piece of the network relies on knowing what time it is—routers, switches, authentication systems. When the time is catastrophically wrong, the whole system loses trust in itself. It's like every component suddenly thinks it's in a different era and stops talking to the others.
And Telstra knew this was a risk?
They did. The government had warned them specifically about it. They understood timing systems were critical. But knowing a risk exists and actually protecting against it are different things. Their backup systems didn't work.
Why would backup systems fail if the company knew the danger?
That's what they're investigating now. It suggests the redundancy they built wasn't tested properly, or wasn't designed to catch this particular failure mode. Jersey's telco had the same vulnerability and learned it the hard way five years earlier.
What troubles you most about this?
That someone may have died because of it. A family tried to call for help and couldn't reach emergency services. Telstra says their records don't show a failed call, but police are investigating. We don't know yet if the outage caused the death, but the fact that we're asking the question at all means the system failed at the moment it mattered most.
Does knowing about a risk beforehand make this worse?
It does. It moves this from "unfortunate technical failure" to "we knew this could happen and didn't prevent it." That's a different kind of accountability.