This is the story of William Woods. No, not the fake William Woods. The real William Woods. It’s a story that traverses more than three decades, zigzagging from a hot dog stand to a jail with a stop at a mental institution and ending with—well, you.

It all began 36 years ago in the sun-drenched streets of Albuquerque, where William Woods was working at a hot dog stand, serving office workers and city dwellers. On an otherwise unremarkable day, his coworker—a large man with dark hair named Matthew Keirans—stole Woods’s wallet. With it, Keirans pilfered not just Woods’s Social Security number but eventually his entire identity. The theft was the seed of an existential usurpation and the beginning of a Kafkaesque nightmare for William Woods. That’s because Keirans had decided to become William Woods to escape his own troubled past. By 1990, Keirans had used the Social Security number to obtain a Colorado ID in Woods’s name; he then opened a bank account (also in Woods’s name) and wrote some checks that later bounced, according to court documents.

This is where reality broke in two. Keirans decided he would straighten out his life, and after stops in Idaho and Oregon, marriage and the birth of a son, he settled in Wisconsin. Within a year, he got a job working in IT for the University of Iowa Hospitals and Clinics, where he made more than $100,00 a year. He bought a house and two cars, and lived a version of the American dream—albeit a slightly different one. He lived all of it as William Woods.

Meanwhile, about 2,000 miles away, the real William Woods’s life was on a very different track. He ended up homeless, living on the unforgiving streets of Los Angeles, doing odd jobs and selling scraps of metal to get by. After almost three decades, he finally discovered what had happened: that Keirans had stolen his identity and was living as him. Woods learned that Keirans maintained deposits at a national bank with a branch in Los Angeles, and had used Woods’s name, Social Security number, and date of birth to accumulate eight loans from credit unions totaling more than $200,000, according to the US Attorney’s Office in the Northern District of Iowa. Seeking to reclaim his identity and not wanting to pay the debt, Woods explained to a branch manager that he was actually William Woods and demanded that the accounts opened in his name be closed. But when the manager called the number on file, Keirans answered and said that the man standing in the bank branch was actually Matthew Keirans, who had stolen his identity, and that the bank should call the police. When the cops showed up, they arrested Woods and held him without bail at the Los Angeles County Jail. The charge: identity theft and false impersonation, both felonies. In other words, he was charged with saying he was William Woods.

Throughout the legal proceedings, Woods maintained that he was William Woods, not Matthew Keirans. But the Los Angeles judge didn’t believe him, and sent Woods to jail for 428 days. When he finally got out and was brought before the judge again, Woods still refused to declare that he was Matthew Keirans; the judge declared him mentally unfit for trial and sent him to a California mental hospital where he was treated with psychotropic drugs and other therapies, and held for 147 days, stuck in the limbo of the psych ward, his reality dismissed as delusion, according to court records. Woods was finally allowed to leave the hospital on one condition: that he plead no contest and admit that he was Matthew Keirans.

In January 2023, Woods found out where Keirans worked and contacted security at the University of Iowa Hospitals and Clinics, which forwarded his concerns to the University of Iowa police. Ian Mallory, a UI police detective, appeared to be the first to believe Woods. He pulled Woods’s birth certificate, then gave his father a DNA test, and then tested Woods, finally proving that William Woods was who he said he was. When Mallory approached Keirans with the irrefutable evidence, he knew it was game over. “My life is over,” he later said, according to court records. Keirans pleaded guilty and was convicted on one count of making a false statement to a National Credit Union Administration insured institution to obtain a loan, and one count of aggravated identity theft. He now faces up to 32 years in prison, a fine that could reach $1.25 million, and five years of supervised release following any imprisonment, according to the US Attorney’s Office in Iowa’s Northern District.

The story of William Woods, real and manipulated, unfolds as a stark reminder of the dangers posed by identity theft, even without employing today's rapidly evolving technology. The advent of advanced AI and deepfake technology has introduced a new kind of threat. An individual’s appearance and voice can be cloned with disturbing accuracy, leading to scenarios in which people are deceived into making financial transactions or disclosing sensitive information.

“Identity theft is getting easier, thanks to technology. It used to be that someone would misuse your credit card or open a fake loan. In the Woods case, his entire life was taken over. That gets easier when AI can generate a huge amount of realistic false documents,” Lou Steinberg, the founder and managing partner of CTM Insights, a cybersecurity research lab, told me, noting that this will only become more prevalent with deepfake technology. “In the Woods case, the fraudster could fool people who didn't know the real victim. In the future, he could have a phone call, Zoom, or FaceTime chat with people who do know the victim and fool those people too.”

There are already numerous stories of AI and deepfake technology being used to steal people’s identity and to perpetrate crimes with the theft of someone’s likeness and, in some instances, entire identity. Earlier this year, a finance worker at a multinational bank was conned into transferring $25 million to criminals during a video conference call, where deepfake technology was used to impersonate the company’s chief financial officer and other staff members. The worker who made the transfer believed the people he was talking to were his coworkers, because they looked and sounded like them. There’s the story of a mom in Georgia getting ready to send hackers $50,000 in ransom money when she was led to believe her daughter had been kidnapped (the hackers stole her daughter’s voice from social media and mimicked it), or the Chinese exchange student who was so frightened by an AI hoax that he ran away from his host family’s home and took his own ransom photos that were used to extort money from his family in China, or the San Francisco couple who believed their son was lying in a pool of blood under a car after being in a terrible accident and forked over $15,500 to a lawyer, who was actually a scammer.

The intersection of Woods’s harrowing personal ordeal under supposedly normal circumstances and the ease with which identities can now be appropriated via technology illuminates a daunting future. Steinberg, the founder of CTM Insights, thinks that we’re about to move past “identity theft”—essentially what Woods experienced on a terrifying longitudinal scale—to an era of “identity hijacking” in which bad actors not only take over your name and claim hold of your history but also use re-creations of your voice, make new images of you, and become a virtual version of you that is indistinguishable from the real you. All, of course, without your authorization.

Just last week, a principal at a Maryland high school was put on leave after a 42-second audio clip surfaced online of him deriding Black students, calling them “ungrateful” and saying that they would be unable to “test their way out of a paper bag.” It was later revealed that the audio was artificially created using AI tools by a disgruntled former athletic director at the school. The athletic director was arrested at Baltimore/Washington International Thurgood Marshall Airport, and faces charges that include stalking, disruption of school operations, and retaliation against a witness, as reported by WBAL TV in Baltimore. The tools used to create the fake AI clip were easily accessible online. Hany Farid, a computer science professor at the University of California, Berkeley, who consulted with police on the case, told The Washington Post that the audio recording had been created using just a few seconds of the principal’s voice.

So, how do we stop this from happening to all of us? The Federal Trade Commission reported recently that it received fraud reports from  2.6 million Americans in 2023, the same as in the previous year, with investment and impostor scams topping the list. Losses from fraud totaled $10 billion. Ben Colman, cofounder and CEO of Reality Defender, which builds deepfake detection software, says the onus should be on the institutions, like banks and schools and government, to ensure we’re not defrauded in the future. “Deepfakes are so convincing that every member of our team—including several with PhDs—have undoubtedly fallen for one at one point. The only way to stop AI is with AI,” Colman said. “The institutions we trust—the same institutions that are at risk of being defrauded—have the responsibility of separating real from fake.”

Clearly, the institutions not only failed Woods, but they were the same institutions that were responsible for throwing him in jail and subsequently a mental hospital. Lucky for Woods, another more old-school technology—DNA—eventually set him free. His case was finally vacated last month by Los Angeles County Superior Court Judge William C. Ryan, who said at the hearing that what happened to Woods was truly astounding. “The word that comes to mind is Kafkaesque,” Ryan said. “Out of the novels of Franz Kafka.” Woods told the Los Angeles Times, “I’m happy, because I knew I was innocent.”