NEW YORK (NYTIMES) - At 2.18pm Eastern on Wednesday (Oct 3), cellphones across the United States emitted the ominous ring of an emergency presidential alert.
It was the first nationwide test of a wireless emergency alert system, designed to warn people of a dire threat, like a terror attack, pandemic or natural disaster.
There is no opting out, which has already prompted a lawsuit.
"THIS IS A TEST of the National Wireless Emergency Alert System," it read. "No action is needed."
Two minutes later, televisions and radios showed test alerts.
There is no notification plan for landlines.
Officials said before the wireless test they believed that it would reach about 75 per cent of the roughly 225 million cellphones in the country, though they hoped the number would be higher. It can take up to 30 minutes for the alerts to be transmitted to all devices.
Some things that could interfere: ongoing phone calls or data transmission, a device that is turned off or out of range, and smaller cellphone providers that are not participating in the programme.
(It is not mandatory, but all the major cellphone providers did take part.)
The test, originally planned for last month but delayed by Hurricane Florence, was the culmination of many years of work. The federal government developed a system to issue the alerts, which are scripted in coordination with numerous government agencies. They are limited to 90 characters, but will be expanded to 360 in the future.
The Communications Act of 1934 gives the president the power to use communications systems in case of an emergency, and a 2006 law called for the Federal Communications Commission to work with the wireless industry to transmit such messages.
The FCC says the resulting Wireless Emergency Alert System has been used by local governments more than 40,000 times since 2012.
The law specified that the system should allow users to unsubscribe from local alerts about extreme weather or missing children.
But the law also states that users cannot opt out of the presidential alerts, which are issued at the direction of the president - or someone he designates, or both - and executed by the Federal Emergency Management Agency.
That has raised concern, given President Donald Trump's proclivity for sending impulsive messages on his favorite platform, Twitter. And it has already spurred a lawsuit, filed in New York last week. The plaintiffs, three New York City residents, say that the alerts violate their free speech and amount to an unconstitutional seizure of their devices.
The system "is tantamount to hijacking private property for the purpose of planting a Government-controlled loudspeaker in the home and on the person of every American," the lawsuit states.
In response, federal officials said that the law laid out strict guidelines for sending emergency alerts. They are sent from secure locations by trained operators using devices similar to laptop computers.
"You would not have the president just waking up and deciding to do it," said Antwane Johnson, who oversees public alerts for FEMA.
After complaints about the impending alert on social media, Verizon sought to reassure customers in a statement posted on its website, saying that events like Hurricane Florence underscore the importance of testing alert systems.
But emergency alerts have come under scrutiny in recent months. In January, the state of Hawaii sent a false alert warning of an incoming ballistic missile, touching off panic across the state. Officials blamed a worker who misunderstood instructions from a supervisor. And during wildfires in California last fall, officials in Northern California were criticized for not sending more alerts.
John Lawson, executive director of the Advanced Warning and Response Network, said that the country's alerting systems were "fragile and fragmented" - and that testing them was crucial to strengthening them.
"I think that the national test is a very good idea," he said. "And I hope that the public will have a little patience with it."