He died in a tragic accident. Why did the internet say he was murdered?

The tracks, where Matthew Sachman, who went by Matteo, fell and was killed by a train in Manhattan. Photo / Natalie Keyssar, The New York Times

Within a day of the death of Matthew Sachman, 19, on New York City subway tracks, so-called obituary pirates had flooded search results with false information.

It was the kind of tragic accident that reverberates through a community: a first-year college student, out late in New York City on New Year’s Eve, falls onto the subway tracks and is killed by an oncoming train.

Word of the 19-year-old’s death spread quickly among the people who knew the young man, Matthew Sachman, who went by Matteo, from his days at Collegiate High School in New York or Georgetown University.

As the circle of people learning the news widened beyond Sachman’s immediate family, concern and curiosity about his shocking death led many of them to Google, where they typed his name and what little they knew into the search bar:

Matteo Sachman subway accident. Matteo Sachman obituary. Matteo Sachman death.

But instead of answers, anybody searching for information was confronted by a blizzard of poorly written news articles, shady-looking YouTube videos and inaccurate obituaries. Some said he was 29 years old and was from Nantucket, in Massachusetts. (His family spent summers there, but he was from New York.)

Others made an even more shocking claim: Sachman, they falsely reported, had been stabbed to death in a Bronx subway station. In fact, according to a statement from his family, Sachman and a friend were fooling around on a platform at the East Broadway stop in Manhattan, when he fell onto the tracks and was killed instantly by a train entering the station.

“There were sites I’d never heard of, information that was completely wrong,” said Peter DeLuca, whose son, Matthew, went to high school with Sachman. “It didn’t make sense.”

And yet the information was spreading. DeLuca, owner of several funeral homes in Manhattan, had come across other dubious obituaries recently and thus sensed something was amiss. But he and his wife were soon hearing from acquaintances who believed Sachman had been stabbed.

The young man’s friends were already grieving. Now many of them were unnerved. In the hours after his death, his name and likeness ricocheted around a dark corner of the internet, where profiteers using artificial intelligence tools capitalised on the anguish and desperation of the people who were mourning him.

“I was looking for the truth,” said Devan Mehrish, 19, a childhood friend, who was scrambling to piece the story together from San Diego. “But I didn’t find it there.”

The ‘pirates’ who hunt for tragedy

In the hours after his death, friends and family scrambled to find out more about Sachman’s death. Few details were available — no obituary, no news stories.

But as people searched Google for information, someone on the other side of the world was searching for exactly the kinds of reverberations that Sachman’s death had caused.

Faisal Shah Khan, an internet marketer in India, knew nothing about Sachman. But suddenly, enough people were searching for “Matteo Sachman” to push his name up a list of trending Google search topics that Khan was monitoring as part of a digital moneymaking scheme.

To Khan, the rising interest meant that an audience for online content that did not yet exist was growing rapidly before his eyes. He was poised to deliver it.

Khan, 30, is part of a booming cottage industry online in which enterprising people take advantage of the void of information in the wake of a sudden tragedy to drive web traffic to hastily assembled articles and YouTube videos.

These so-called obituary pirates seem to know about the deaths of everyday Americans long before they have been reported publicly anywhere else.

Khan — whose website, FSK Hub, was the first site identified by The New York Times to post anything about Sachman’s death — agreed to walk the Times through his process.

Khan has spent the past five years building an online advertising business with websites devoted to celebrity news and tech reviews. But he said obituaries make up a huge part of his content farm.

Working from his living room in New Delhi, he closely monitors Google Trends for activity related to certain grim keywords: obituary, accident, death.

Google allows anyone to track usage trends for search terms in windows of time as narrow as the previous hour. When Khan searches those keywords on Google Trends, the company shows what else people who are searching for those terms are actively searching for in the moment.

Matteo Sachman subway accident. Matteo Sachman obituary. Matteo Sachman death.

These were the kinds of searches for truth that precipitated the flow of misinformation back to the people doing the searching.

Based on related searches, such as “subway accident,” Khan could surmise how Sachman had died. Khan could then conduct a cursory search of his own around the internet for any biographical information, leading him to a LinkedIn page detailing Sachman’s work history. And finally, he could prompt an AI tool called a large language model to create a short article.

“The article should be written in a conversational style, using personal pronouns, rhetorical questions and analogies to engage the reader,” read one prompt intended for a language model, which was accidentally published on FSK Hub.

For his effort, Khan said he earns thousands of dollars each month from Google’s ad network using numerous websites. But experts in internet marketing suggested that a more likely estimate would be less than $100 per month. Articles such as Sachman’s obituary would only fetch a penny or two per month, experts said.

“Money is not my priority,” said Khan in an interview. He described himself as an “avid equestrian” who unwinds by riding horses through the wilderness and dreams of becoming a journalist one day. “Writing blogs brings me joy,” he added.

In all, it took barely more than 24 hours for the search results of Sachman’s name to be transformed into a minefield of ill-conceived, misleading content.

“We were trying to find out what happened, but we saw some weird things,” said David Lombardi, owner of a nursery and furniture store in Nantucket where Sachman had worked a summer job. “I just stopped and thought, ‘This doesn’t feel right.’”

A growing problem

All anyone wants, in the event of a loved one’s death, is for the memory of the deceased to be treated with dignity. The machine-generated obituaries are the opposite of that.

What the Sachman family experienced is becoming increasingly common for the bereaved, according to scores of complaints published online and interviews with people who endured eerily similar ordeals.

They were alarmed to see the fake obituaries posted before any official announcements or news stories. They were upset to read baseless speculation about possible causes of death in the AI-generated articles. And they were disheartened that seemingly nothing was being done about it.

“It’s horrific and disturbing because it’s so predatory,” said Chris Silver Smith, a digital marketing consultant in Austin, Texas, whose brother-in-law was killed in a car accident in September.

People believe they can rely on Google not only for news about a person’s death but also for information about funeral services or to donate money to a grieving family.

Yet many now step away feeling unsettled or angry or, in some cases, fearful that they had been exposed to digital scams or malware.

“I felt like I had just clicked on something that was going to make my phone blow up,” said Peggy Hammond, 54, a teacher in Birmingham, Alabama, who began deleting applications, in a panic, after seeing several sketchy articles this month about an acquaintance’s death. “It makes you not trust Google.”

For now, there seem to be few options for anyone hoping to have these fake obituaries removed.

Audrey Wade, 27, a graduate student in Salt Lake City, filed dozens of complaints to Google in August about websites and YouTube videos that targeted a friend who had unexpectedly died. But she never heard back, and the content was never removed.

“They created this platform, with all of its unanticipated possibilities, and are therefore responsible for managing its use and abuse better,” she said.

Although the obituaries appeared on other search engines, including Microsoft Bing, Google played a unique role in spreading news about Sachman across the internet.

Google Trends surfaced terms that internet marketers such as Khan could use. Google Search circulated the obituaries for Sachman’s family and friends to find. YouTube, which is owned by Google, allowed creators to spread falsehoods about Sachman to different audiences. And ads powered by Google appeared on most of the obituary websites, earning revenue for the websites’ owners — and for Google.

A Google spokesperson said the company is aware that low-quality obituaries can flood search results after someone dies, and it is exploring ways to address it.

The company pointed to policies that penalise websites for using automation, including AI, and said an upcoming change to Google Trends will reduce the chances that names of people who are not well known will be highlighted.

“Google always aims to surface high quality information, but data voids are a known challenge for all search engines,” a Google spokesperson shared in an emailed statement. “We understand how distressing this content can be, and we’re working to launch updates that will significantly improve search results for queries like these.”

Google removed several YouTube videos about Sachman after the Times inquired about them and disabled ads on many of the obituary websites.

After the pirates move on

Sachman made the most of his lone semester at Georgetown. He joined the university’s marketing association, volunteered with an organisation aiding homeless people and worked at a bar on campus, according to The Hoya, the school’s newspaper, which on Jan. 19 published the first legitimate news story about his death.

“What happened is devastating, and it hit the community hard, and we miss him,” said Mehrish, who described his childhood friend as charismatic and smart. “It’s been a hard pill to swallow. We haven’t experienced something like this before. It’s a novel experience, and it’s a brutal experience.”

Collegiate, the elite private boys school on the Upper West Side that Sachman had attended, sent out an email on the night of January 2 noting that he had died “as a result of a tragic accident.” Georgetown did the same, with the same vague language and no other details.

The family had wanted to divulge information at their own pace. But in the depths of their grief, as they came together at home to process the excruciating pain of losing their son, Sachman’s parents were already hearing from family members that obituaries with false information were proliferating online — the wrong age, the wrong location, the wrong everything.

“The stories published online were completely wrong,” the family said in a statement.

On January 5, four days after their son’s death, the Sachmans had a paid obituary published in the Times, remembering their son as “a young man full of joy who exuded warmth and love, positive energy and care for all.”

But by then, the clamour for information had passed. The erroneous stories about Sachman remained, though no longer at the very top of the search results for his name.

The pirates who published them, having extracted a few pennies, had long since moved on.

This article originally appeared in The New York Times.

Written by: Andrew Keh and Stuart A. Thompson

Photographs by: Natalie Keyssar

©2024 THE NEW YORK TIMES