Day 6

How the internet broke the emergency response system

False alarms in Hawaii and Japan have put emergency alert systems under scrutiny, just as Canada gets ready to adopt our own text message alert system.
Cars drive past a highway sign cancelling a missile alert on the H-1 Freeway in Honolulu. (Anthony Quintano/Associated Press)

Hawaiians were subjected to a surreal experience last Saturday morning when they awoke to a government alert about an inbound ballistic missile attack.

This smartphone screen capture shows an erroneous incoming ballistic missile emergency alert sent from the Hawaii Emergency Management Agency system on Jan. 13, 2018. (Caleb Jones/Associated Press)
"THIS IS NOT A DRILL," the smartphone notification declared in all-caps. The television and radio alerts were similarly ominous, leaving local residents terrified and desperate to get in touch with loved ones and find a safe place to hide.

We know now that the panic-inducing alert was a false alarm, sent in error by the Hawaii Emergency Management Agency during a regular system drill. 

Then, on Tuesday, it happened again — this time in Japan, where the public broadcaster NHK errantly issued a similar warning about a North Korean missile launch on its website and mobile app.

The headline-grabbing errors left people around the world wondering how they'd find out if a missile was actually headed their way.

In Canada, the government still relies almost exclusively on TV and radio alerts — although the CRTC plans to have an emergency text message alert system in place by April.

But this week's false alarms have illustrated that the new systems are anything but foolproof — and the Internet may be to blame, according to Ian Bogost, a contributing editor with The Atlantic and professor of interactive computing at the Georgia Institute of Technology.



From sirens to SMS

Long before digital technology made it possible to instantaneously send a message to millions of smartphones at once, emergency alert systems — from air raid sirens during World War Two to the televised nuclear missile drills of the Cold War — were implemented to warn the public of potential peril.

"The purpose of these things was to warn citizens about an imminent threat, typically of nuclear war, and then to provide additional information about the state of the government," Bogost tells Day 6 host Brent Bambury.

"If you imagine that Washington, D.C. was attacked and perhaps key personnel in the government had been killed, the purpose of the system in part is to notify the populace that the government is still running and how it's doing so."

Back then, however, false alarms were generally limited to the area where the alert was being broadcast.

"By and large, this kind of national or international failure was much less common, because it almost couldn't happen," Bogost says.

When you get a wireless alert, what do you do? Where do you go?- Ian Bogost , contributing editor with The Atlantic

"These systems were designed to broadcast emergencies very, very locally. And so a failure or an error would have had less of an impact than it does today, because everyone with their smartphone receives this thing, and then they immediately start tweeting and posting to Facebook and communicating with everyone else all around the world about it."

In the pre-digital era, emergency alerts were effective in part because they served as a means to an end, Bogost says.

"In the past ... you see the emergency alert and then moments later some kind of a news program would come on and say, 'Here's some more information — here's what you need to know'," he explains.

"What the alert was meant to do was just to get your attention — 'Hey, there's something important that you need to know'. But when you get a wireless alert, what do you do? Where do you go?"


Local residents take part in an emergency drill in the wake of repeated missile launches by North Korea on Japan's northernmost main island of Hokkaido on Sept. 1, 2017. (Reuters)

Digital dilemma

In Japan, the erroneous alert was corrected within just five minutes — but in Hawaii, it took 38 minutes for officials to issue a second alert telling residents there was no threat. 

That left plenty of time for panic to take hold as locals shared the alert on social media outlets like Facebook and Twitter, drawing international media attention to the story.

A highway median sign broadcasting a cancelled threat alert in Kaneohe, Hawaii after emergency officials mistakenly sent out an alert warning of an imminent missile strike on Jan. 13, 2018. (Jhune Liwanag/Associated Press)
The efficiency of the alert's spread may have caused more harm than good, according to Bogost.

"It's useful to reflect on how this false alarm would have come to the public knowledge in the 1970s and 1980s," he says. "What probably would have happened is that the whole thing would've gone down in Hawaii, and then on the nightly news all across the world, we would have heard about it that evening," he notes.

"But it wouldn't have raised a kind of national and global panic immediately, with everyone demanding answers and calling for the head of whoever sent the message falsely. That's certainly something the internet has brought about, and [it] may be incompatible with emergency notifications that give citizens good information about what to do next."


This smartphone screen capture shows a false incoming ballistic missile emergency alert sent from the Hawaii Emergency Management Agency system on Saturday, Jan. 13, 2018. (AP Photo/Jennifer Kelleher) (Jennifer Kelleher/The Associated Press)

Improving alerts

Part of the issue with the text message alerts that were sent in Hawaii was the character-count limit, which made it difficult for the EMA to convey enough useful information, Bogost says.

One of the biggest things is just making sure that additional information is available and people know how to get it.- Ian Bogost

"There are proposals to add more information to the messages themselves — right now, they're limited to 90 characters," he says. "And maybe if it included a URL, which is one of the proposals, then you could follow a link somewhere."

"But the problem with that is that often in a case of an emergency when everyone tries to visit the same website, the server can't handle it and then creates the same panic."

Bogost says governments around the world will need to take a hard look at the recent false alarms to better understand how to tweak their own emergency alert systems to ensure they're both adequate and accurate in the wake of a true threat.

"Certainly it's important to learn from what's gone wrong all around the world, in Japan and in the United States," he says. "And I think one of the biggest things is just making sure that additional information is available and people know how to get it."


To hear the full interview with Ian Bogost, download our podcast or click the 'Listen' button at the top of this page.