EVERETT — The detectives came knocking at her Tumwater home on a warm evening in spring.
Chelsea Rustad wasn’t in trouble, but police wanted to chat about one of her relatives. He was under investigation for murder.
Maybe, she thought, they had the wrong house.
“Your mind races,” she said. “Is it an immediate family member? Where is this going? They’re kind of seeing the confusion and the apprehension on my face. Why would you be on my door step, coming from Skagit County?”
That same day in 2018, police arrested her second-cousin, William Talbott II. Rustad knew the name but not the man, except as a node on her family tree on Ancestry.com.
Last week, a Snohomish County judge sentenced Talbott to two life terms in prison for murdering a young Canadian couple in late 1987.
For decades sheriff’s detectives had searched for the person who shot Tanya Van Cuylenborg, 18, and strangled Jay Cook, 20.
Without knowing it, Rustad provided detectives the critical clue to solve the killings: her family’s DNA.
Police never came across Talbott’s name until last year, when a private lab, Parabon, extracted DNA from semen found on Van Cuylenborg’s pants. The lab uploaded the profile to GEDMatch, a site where people can import data from genetic tests conducted by ancestry companies like 23andMe, AncestryDNA and others, to find relatives.
A genealogist worked backward to identify the killer by building branches on both sides of his family tree.
Rustad, a quality assurance technician at the Department of Fish & Wildlife, is Talbott’s grandmother’s brother’s granddaughter.
Privacy advocates cried foul when they learned police around the country were using genealogy banks to trawl for suspects — sifting through citizens’ genetic data without a warrant, without notice, and without oversight.
Rustad, too, has long held her own misgivings about law enforcement, both personal and political.
Yet she hardly could’ve been happier, for the sake of the families of the two victims, when she learned her DNA led police to a killer.
In fall 2015, a contest popped up on Rustad’s Facebook feed asking people to share their favorite Halloween memory.
The prize: a free test from AncestryDNA.
Rustad posted a photo of herself in a ballerina costume around age 7. More than 300 people entered. Rustad won. A couple of weeks later, a spit kit arrived in the mail.
Inside the box was a note: “Your DNA has a story. It’s time to discover it.”
One of her cousins mentioned that she should upload her profile to GEDMatch. So she did. The third-party site can help people find blood relatives, who happened to test through a different company.
Rustad imported her data, fiddled around on the site, and promptly forgot about it.
She’d first started figuring out her family tree on Ancestry about six years ago, when she realized she couldn’t name off her great-grandparents. For her, the most exciting part was compiling portraits, and seeing the faces of people who share her genes.
The further back you go, the less reliable peoples’ Ancestry trees become. At about the great-great-grandparent level, it often turns into a giant game of telephone. People copy each other’s trees, and there are precious few documents. People can claim to be descended from royalty, but without sources to back it up, that’s often a red herring.
Rustad pieced together the Talbott branch of her family tree about three years ago, as she fleshed out descendants of her great-aunt Blanche. She’d never met Talbott or his sisters. All of his siblings were mentioned by name in about a dozen sources on Ancestry — yearbook photos, Census records, and so on. William II, it seemed, was the odd man out. He’d left a small footprint in public records, perhaps three mentions.
“He was very off the radar,” she said. “Because he never married or had kids, there was no marriage record or something of that nature, which makes it harder to learn about a person.”
Rustad befriended Talbott’s sisters on social media. They shared family photos, but never anything recent of Talbott.
“He’d been a person that nobody spoke about,” she said.
On her doorstep, when detectives mentioned the suspect’s name, Rustad thought they were talking about his father, William Talbott Sr.
“I had no reason to think, ‘Oh, maybe they’re talking about such-and-such,’ because that’s not something that has ever happened in my family, a violent crime — let alone murder,” she said.
By then, it had literally been years since Rustad thought about GEDMatch. Her fiancé grabbed his laptop. They logged on to walk police through her family history, and how she knew for certain that she and Talbott were second cousins. Rustad had only hoped to solve her own family mysteries. Instead, by having retraced her family tree, she’d eliminated much of the legwork for the genetic researcher working with authorities.
Genetic genealogy had provided an investigative lead, but like any good tip, detectives still needed to fill in the blanks.
Undercover officers had been following Talbott for days as he drove his semi-trailer on routes to south Seattle, until he opened the door and a paper cup tumbled into the street. A test on DNA from the recovered cup confirmed the genealogy work was correct: The crime scene DNA came from Talbott. Detectives would learn he’d grown up just seven miles west of a rural bridge over the Snoqualmie River, where Cook’s body was found.
The day after the visit from police, Rustad saw the news that authorities had the SeaTac trucker in custody.
Talbott, 56, had no felony record. His genetic profile was never in any federal law enforcement database.
Police only found him because two of his distant cousins went looking for their own pasts.
To get a meaningful match, generally you need to track down a second cousin or closer. Exactly one second cousin on each side of Talbott’s family had shared data on GEDMatch. The families intersected in the marriage of a Woodinville couple, who had one son.
“After what (the victims’ families) went through, it’s about time they got good luck to come their way,” Rustad said. “It’s a shame they had to wait so long for the technology to catch up.”
Twice in her life, Rustad has called police to report a serious crime against her.
Both were assaults by people she knew.
Both times there was little, if any, physical evidence to back up her stories. In both instances she felt the justice system failed her miserably.
“Everyone blames you, if you are the one who’s abused,” Rustad said. “You are the one on trial.”
In the first case, she told police a supposed friend threatened her with a shotgun.
The second time, the cops hardly investigated a serious felony assault by a different man, she said.
“They sure did nothing about it, other than making my life worse,” Rustad said.
In her personal politics, she sees police as a tool for mass incarceration and institutional oppression. Some of the earliest forms of law enforcement in America, she said, were slave patrols in the South. She believes police can get away with “darn near anything with impunity.” She points to systemic injustices, like the fact that police departments across the country have a backlog of rape kits that have sat on shelves, untested, for years. Each of those represents a person, she said.
Around the time of the 2016 election, Rustad was feeling pretty cynical about her country’s leaders. So she ran for Tumwater City Council as a socialist.
She lost. In a two-way race, she took about 26 percent of the vote — better than she expected, given her pledge to avoid money from capitalists.
“I believe in doing things for the right reasons,” she said, “regardless of what the outcome is going to be.”
In spite of all her skepticism of authority, Rustad is glad police used her genetic data, even without asking. She does not see it as a breach of trust.
“It’s not a bad thing a killer was apprehended, and a suspect was finally found, and the victims’ families finally have answers,” she said. “These are all people who have opted to share DNA on this site.”
Public outcry over genetic companies quietly cooperating with law enforcement was great enough for GEDMatch to change its policy this year, restricting police access unless users checked a box to “opt in” and say their genes can be used to fight crime.
The policy changes show the people behind the site are listening, Rustad said. But the way she sees it, she chose to share her data with the public in the first place.
“It would be like saying if I found a note by Talbott on the ground saying, ‘Why I did it,’” she said. “It was laying there, and I found it. If any random person could sign up for a website and see public content, it’s pretty available (to police).”
And in this case, genetic genealogy was the only way to move forward.
“I understand the privacy concern,” Rustad said. “But I look at the outcome of what came of it, and I’m glad. … I wish everyone could have their situation treated with the same respect and seriousness.”
Rustad kept a close eye on Talbott’s case, even when media reported the smallest details.
The two Canadians disappeared on what was to supposed to be an overnight trip from Vancouver Island to Seattle, to pick up about $750 in furnace parts for a family business.
Six days later, on Nov. 24, 1987, a passerby found the young woman’s body near a roadside drainage pipe south of Alger. Two days later, a hunter uncovered Cook’s body south of Monroe.
For their loved ones, 30-plus years of painful waiting followed.
Because of genetic genealogy, in the past year long-awaited breakthroughs arrived for dozens of other cold investigations nationwide.
The first high-profile arrest came in April 2018 in the case of the Golden State Killer. Days later, Talbott was in jail. On June 28, he became the first person in the country to be convicted by a jury in a case cracked through ancestry research.
As far as she knew, Rustad was the only member of Talbott’s family in the courtroom. She wanted to support the victims’ families. She’d been in touch with Cook’s brother-in-law, but now met in person. Rustad carried a bouquet of flowers. She sat in the gallery next to Tanya’s brother, John Van Cuylenborg.
No one spoke on the defendant’s behalf, except for a defense attorney and Talbott himself. Talbott told the judge he was innocent and could not even comprehend the level of violence in the two murders. His voice cracked with emotion.
“I am someone convicted of a crime that I did not commit,” he said.
Talbott hadn’t testified at his trial, remaining silent through the proceedings.
When he spoke Wednesday, Rustad did not believe him.
“You can say you’re innocent,” she said, “but if you’re not going to offer your version of events, we’re forced to look at what we do know. And the evidence doesn’t align with that. It really negates it, completely.”
Though he is kin, Rustad feels no sympathy for Talbott.
By consenting to let police use her genetic data on GEDMatch, it opened the door to possibly implicate anybody who was a second cousin or closer: parents, sisters, brothers, sons, grandmothers, and so on.
That’s OK with Rustad.
Her family has been behind her decision, she said.
“If anything,” she said, “(Talbott) is the one that has brought his family into something they didn’t sign up for, by going out and committing these crimes. … And he harmed people. He brought (the victims) into something they never signed up for.”
In an interview with a reporter this month, Rustad realized she hadn’t been on GEDMatch since the site changed its policy requiring customer consent. By default, her DNA had been unavailable to law enforcement for months.
So for the first time in forever, she logged back into the site, to check the box.
Caleb Hutton: 425-339-3454; firstname.lastname@example.org. Twitter: @snocaleb.