Customers trying to get a vehicle title in Columbiana County on Friday saw this sign on the door, with the Lisbon office falling victim to the nationwide Microsoft system glitch that has affected airports, Bureau of Motor Vehicle offices, title offices and other businesses that use the system. Shane Patrone, Columbiana County’s Chief Deputy Clerk of the Clerk of Courts office, said the title office remained open, with staff offering to take and process paperwork when the system is up and running again. Passports were able to be processed. Patrone noted that the rest of the county’s systems through the Clerk of Courts were working fine, with the courts and other offices having no issues. He said the title office is connected to a state system that has been down across the state. (Photo by Mary Ann Greier)

NEW YORK (AP) — A faulty software update caused technological chaos around the world Friday, grounding planes, shutting down some financial firms and news organizations and disrupting hospitals, small businesses and government offices.

The scale of the outages underscored the fragility of a digitalized world that relies on just a few providers for essential computer services.

The problem was caused by an update released by cybersecurity firm CrowdStrike and affected only customers running Microsoft Windows, the world’s most popular personal computer operating system. It was not the result of a hack or cyberattack, according to CrowdStrike, which apologized and said a fix was on the way.

Businesses and governments around the world experienced hours of disruption — their computer screens glowed blue with error messages — and they scrambled to fix the fallout. CrowdStrike’s CEO said some of its systems will require time-consuming manual fixes.

Thousands of flights were canceled and tens of thousands were delayed, leading to long lines at airports in the U.S., Europe, Asia and Latin America. Airlines lost access to check-in and booking services in the middle of the summer season. By late afternoon Eastern Time, the worst appeared to be over, although cancellations and delays continued as the disruption cascaded.

Several local U.S. television stations were unable to broadcast news Friday morning, and some state and local governments reported problems at courts, motor vehicle departments, unemployment offices, 911 centers and other offices. But most systems were back to normal as the day progressed.

The affected hospitals had problems with their appointment systems, forcing them to suspend patient visits and cancel some surgeries.

Alison Baulos said her 73-year-old father’s heart surgery in Paducah, Kentucky, was canceled Friday morning because of the outage, leaving her family fearful and worried.

“It really makes you realize how dependent we are on technology and how scary it is,” Baulos said in an interview. She said her father was waiting at Baptist Hospital to hear what would happen next. A call to the hospital was not immediately returned.

American Express said there were temporary issues processing transactions, while TD Bank responded to online complaints by saying it was working to get customers back into their accounts.

Elsewhere, people experienced minor inconveniences, such as problems with pre-ordering at Starbucks, which led to long lines at some of the coffee chain’s locations.

In Times Square in New York City, just before 12:30 a.m., the blue “recovery” screens that appeared on laptops appeared on several giant electronic billboards. A few were dark on Friday afternoon.

A disturbing memory

of vulnerability

According to cyber expert James Bore, real damage would occur. “All these systems run the same software,” said Bore. “We’ve made all of these tools so pervasive that when something inevitably goes wrong – and it will, as we’ve seen – it will go wrong at scale.”

The head of the German IT Security Agency, Claudia Plattner, said “We cannot expect a solution to come quickly.” It is difficult to make a prediction about exactly when all systems will be operational, but “it won’t take hours”, she added.

CrowdStrike said in a recording on its customer service line that the issue was related to “the Falcon sensor,” referring to one of its products used to block online attacks. The company says it has 29,000 customers.

In an interview on NBC’s “Today Show,” CrowdStrike CEO George Kurtz apologized and said the company… “We are deeply sorry for the impact we have caused to customers, travelers and everyone affected, including our businesses.”

“We know what the problem is” and are working to fix it, Kurtz said. He noted, however, that it could take some time “some time” for certain customers, especially those who do not have the necessary expertise themselves.

While CrowdStrike’s update was automated, the fix requires manual work, such as deleting corrupted files, which can take days or longer for some customers, according to Forrester analyst Allie Mellen.

“Given that CrowdStrike has a lot of customers, a lot of Fortune 500 customers, and they probably have millions of (computers) under management, this creates a bigger problem,” said Mellen. “It will be a long and difficult process.”

Ann Johnson, corporate vice president and deputy chief information security officer at Microsoft, said late Friday afternoon that “At this point I would say that customers are receiving or have received the information they need and are getting the support they need. They understand that it is a very big problem.”

She said Microsoft’s main focus is on getting customers back online, but she couldn’t estimate how long that might take.

In Alaska, the state’s court system is back up and running after 12 hours of repairs, spokeswoman Rebecca Koford said. In Iowa, Gov. Kim Reynolds said the state’s critical technology systems were back up and running by mid-afternoon.

Shares of CrowdStrike, based in Austin, Texas, closed down more than 11% on Friday. Microsoft’s stock fell less than 1%.

Although the impact of the outage was felt everywhere, forecasting agency Capital Economics said it was likely to have little impact on the global economy.

Cybersecurity experts say people affected by the outage should also beware of malicious actors claiming to be able to help. “Attackers will definitely target organizations because of this,” according to Gartner analyst Eric Grenier.

Kurtz wrote in a letter to customers on CrowdStrike’s website that the outage did not impact Falcon systems or security scans.

Air travel delayed everywhere

Most airlines blamed the problems on their booking systems. Thousands of flights were affected in the U.S. alone, but by late morning, East Coast carriers said they were beginning to ease the problems and resume some service. Unclogging the system, however, takes time.

At Minneapolis-Saint Paul International Airport, Sarah Schafer was delayed at her cousin’s 50th birthday party in Florida. She had been waiting for nearly three hours with no indication of when her flight would be rebooked.

“I seem calm,” said Schafer, who used a cane because of an ankle injury. “But my angry side might come out.”

Airlines and railways in the UK faced long waits, and airports across Europe suspended landings or halted take-offs for several hours due to problems checking in passengers.

Saskia Oettinghaus, a member of the German Olympic diving team, was one of those stranded at Berlin airport.

“We are on our way to Paris for the Olympic Games and now we are standing still here for the time being,” said Oettinghaus.

In Cancun, Mexico, the top tourist destination on the Caribbean coast, the state government reported 24 cancellations and 100 delayed flights. Some travelers tried to cheer up the long waits by singing the traditional Mexican song, “Silito Lindo,” while a band that was also stranded played at Cancun airport.

Broadcasters go silent, operations are postponed, ‘blue screens of death’

In Australia, national news channels — including the ABC and Sky News Australia — were unable to broadcast for hours. Some newsreaders went on air from dark offices, facing computers with blue error screens.

In the U.S., KSHB-TV in Kansas City, Missouri, aired Scripps News instead of local news until about 5:35 a.m., the stations reported on their website. Other local Scripps-owned stations reported similar problems, though Scripps spokesman Michael Perry said early Friday that 90% of the stations were able to air local news.

Hospitals in several countries also reported problems.

According to the UK’s National Health Service, the outage caused problems at most doctors’ surgeries, as appointment and patient record systems were affected.

At Mass General Brigham, the largest health care system in Massachusetts, all scheduled non-urgent surgeries, procedures and medical visits were canceled Friday because of the outage, a spokesperson said.

International shipping was also affected.

A major container hub in the Baltic port of Gdansk, Poland, said it was experiencing problems. And at the twin ports of Los Angeles and Long Beach, sea terminals were hit, although the outage did not cause significant disruptions.

Today’s latest news and more in your inbox