Google’s failed alert system during the 2023 Türkiye-Syria earthquake raises urgent questions about overreliance on private tech firms in life-or-death moments
It was just after 4 a.m. on February 6, 2023, when the ground began to convulse beneath southern Türkiye and northern Syria. In a matter of minutes, entire apartment blocks collapsed. Cities crumbled. Families vanished.
The death toll soared past 59,000. And Google’s earthquake warning system — designed to give people a few seconds to prepare — stayed largely silent.
A promise undone by silence
Türkiye had adopted Google’s early warning technology to provide real-time alerts to its citizens. The system works by detecting seismic waves through smartphones and issuing mass notifications before the shaking begins.
That morning, almost no one got the alert.
Google later admitted the failure, citing an underestimation of the quake’s magnitude. But the damage wasn’t just physical. It was a blow to public trust in a system that had been billed as the next frontier in disaster readiness.
And that’s what makes this failure so disturbing: it wasn’t a one-off glitch. It exposed a deeper, more dangerous reality — that we’ve handed over a critical part of emergency infrastructure to corporations we barely hold accountable.
The warning that never came
The aftermath in Türkiye was chaotic. Survivors searched for loved ones under rubble. Mobile networks were overloaded. Internet access sputtered out in several provinces.
And in the middle of this collapse, people waited for help — or any kind of official message.
One aid worker in Hatay described it plainly: “Phones didn’t ring. No messages. People just started screaming when the walls moved.”
For a system built on speed and scale, Google’s earthquake alert didn’t just fail technically. It failed conceptually — by assuming that connectivity and computation would hold up under duress.
When infrastructure fails, tech does too
Digital systems are powerful — until the grid goes down.
Disasters like earthquakes, tsunamis, and hurricanes don’t just break buildings. They sever power lines, topple cell towers, and fry fiber-optic cables. When the backbone of modern communication crumbles, digital alerts can vanish into thin air.
That’s not theoretical.
-
In the Türkiye quake, mobile towers were disabled in Gaziantep and Adana
-
During Hurricane Ida (2021), nearly 1 million people lost connectivity in Louisiana
-
In Nepal’s 2015 quake, early warning systems couldn’t reach rural zones due to downed infrastructure
In each case, technology was available. But it was useless without the pipes to deliver it.
The risk of blind algorithms and unequal reach
There’s another problem, and it’s one we don’t talk about enough: who gets the warning — and who doesn’t.
Tech-driven systems often rely on AI, big data, and predictive modeling to decide who’s at risk. But these tools can’t think in human terms. They don’t see fragile housing, old age, language barriers, or rural isolation.
That’s how entire communities get overlooked.
In Türkiye, informal settlements on the urban outskirts received no warning. Elderly residents without smartphones were left in the dark. Some apps didn’t support Turkish dialects. Others didn’t refresh in time.
One paragraph in a government report put it bluntly: “Algorithmic gaps contributed to disproportionate harm in lower-income neighborhoods.”
The consequences of these blind spots can be fatal — especially when seconds matter.
Who gets left behind?
Let’s break it down. Not everyone has an iPhone. Or reliable 5G. Or the digital literacy to navigate emergency apps.
Here’s where the digital divide becomes deadly:
-
Urban vs Rural: Cities get better coverage and faster alerts. Villages get slower signals — if any.
-
Rich vs Poor: Premium smartphones support the latest tech. Older models don’t.
-
Young vs Old: Younger people understand apps. Older folks still rely on radio or sirens.
-
Native speakers vs Immigrants: Many systems don’t account for multilingual communities.
And yet, we continue to design disaster systems that assume universal access — when the reality couldn’t be more fragmented.
One system, too much power
There’s a deeper issue lurking beneath all of this: accountability. Or rather, the lack of it.
Private tech companies like Google now control the levers of public safety in many regions. Their tools shape who gets alerts, how fast they get help, and where resources go.
But these systems are proprietary. They’re black boxes. The public doesn’t get to see how decisions are made, how data is used, or what failsafes exist.
There’s no hotline for tech failures during disasters. No ombudsman. No elected oversight.
When Google’s alert failed in Türkiye, no public investigation followed. No parliamentary hearings. No binding reforms. Just a quiet admission — and business as usual.
The data dilemma: surveillance vs safety
To make real-time alerts work, companies need data. Lots of it. GPS locations. Device movement. Communication logs.
That data can save lives. But it can also be abused.
Without strict regulations, sensitive information collected during crises could be repurposed — for advertising, profiling, or worse.
People deserve to know:
-
What data is being collected during a disaster
-
How long it’s kept
-
Who it’s shared with
-
What legal rights they have
Right now, in most countries, those answers don’t exist.
Can we do better? Actually, yes
The Türkiye quake wasn’t just a wake-up call. It was a blaring siren. And the lesson is clear: digital tools must support — not replace — public systems.
Siren towers. Community leaders. Radio broadcasts. School drills. These old-school methods still save lives. They don’t rely on battery life, signal strength, or cloud algorithms.
Digital alerts should augment, not dominate.
There’s a role for tech — a big one. But it has to be transparent, inclusive, and backed by human oversight.
Governments should set open standards for alerts, fund local training, and ensure every system works offline as well as online.
When the next quake hits — and it will — we need to be ready. With everything we’ve got.
