THE SETUP. In 2014 a 30-year-old woman became the youngest self-made female billionaire in the world. Her company said it could run hundreds of medical tests from a few drops of blood — no needles, no tubes, results in hours, a fraction of the price.
The board had two former Secretaries of State, a former Secretary of Defense, and a man who would become one. Investors included Rupert Murdoch, the Walton family, and the family of a future Secretary of Education. At its peak the company was valued at $9 billion.
There was one problem. The technology didn't work, and the people closest to it had been saying so for years.
WHY IT WAS HARD. This is the part most tellings skip, and it's the part that makes the fraud legible. A blood test isn't one thing. Counting red and white cells requires the cells intact. Checking cholesterol or glucose means mixing blood with chemicals and measuring a color change. Hormones and infections need antibodies that latch onto one specific molecule. Hospital labs run those on separate machines, some the size of a refrigerator.
Theranos promised all of it, in one box, from a few drops. And less blood creates its own physics problem: to stretch a tiny sample you dilute it, and the more you dilute it, the fainter the signal gets, until the machine can't tell a real result from noise.
That's not one invention. That's several breakthroughs at once, inside one small box.
THE FIRST PERSON WHO SAID NO. In 2006 the company's CFO, Henry Mosley, learned that results shown to Novartis had been made to look better than they were. He told Holmes they couldn't keep doing it. According to John Carreyrou's reporting, she told him he wasn't a team player and that he should leave right then.
Theranos never hired another CFO.
Two years later the board decided she was too young and inexperienced to run the company. They called her in to tell her. Two hours later she was still CEO.
THE CHAIN. So how does a company that keeps losing its skeptics keep finding believers? It starts with one man. In July 2011, George Shultz — former Secretary of State — sat down with Holmes for what was meant to be a short meeting. It ran two and a half hours. He joined the board.
Once Shultz was in, every other name got easier. He brought his friends: another former Secretary of State, a former Secretary of Defense, two former senators, a four-star general, an admiral. Not one of them had ever run a medical lab. They didn't need to. Each name made the next one feel safe.
That board became the proof, and the proof brought the money — nine-figure checks from some of the richest families in America, written not because anyone had seen the machine work, but because they'd seen who else was in the room.
WALGREENS. The pharmacy chain wanted Theranos centers in its stores, and hired a lab consultant named Kevin Hunter to check the technology. Hunter asked to see the lab. Theranos said no. He proposed a side-by-side comparison with a regular lab. Theranos said no. They drew blood from a senior Walgreens executive, and the results never came back.
Hunter raised the alarm. The executive running the project overruled him with the sentence that explains the entire deal:
"We can't not pursue this. We can't risk a scenario where CVS has a deal with them in 6 months and it ends up being real."
Fear of missing out beat the evidence. In 2013 Theranos centers opened inside Walgreens stores in Arizona. Safeway went further and spent $350 million building clinics in more than 800 stores. One of the reports that helped win Walgreens over carried the logos of Pfizer and Schering-Plough. Years later, Holmes admitted on the stand that she added those logos herself.
And once it's in your local pharmacy, you assume someone, somewhere, checked.
WHAT WAS ACTUALLY HAPPENING. Follow one sample. You walk into a Walgreens in Phoenix, they prick your finger, and it ships to the Theranos lab in California. That's where the story splits.
For most tests, your blood never touched an Edison. It went to another room full of ordinary analyzers Theranos bought from Siemens and others — machines built for tubes of blood from your arm, not a few drops from your finger. So staff diluted the samples to give the machines enough liquid to read. Which brings the signal problem right back.
By the end of 2014, the Edison was running just 15 of roughly 240 tests Theranos offered, according to the Wall Street Journal. Prosecutors later said that even on those, it failed more than half the time.
And the demos? Visitors got their finger pricked, watched the cartridge slide in, and saw the lights come on. Former employees testified the demos ran on special software that hid error messages, and that the machine could run a null protocol — it looked busy and tested nothing.
In July 2015 Vice President Joe Biden toured the lab. According to the Journal, Theranos didn't show him the real one. They cleared a room, repainted it, and lined the walls with metal shelves stacked with their newest machines. Biden called it "the laboratory of the future."
THE SILOS. The secret held because almost nobody could see the whole thing. Teams were kept apart; the people building the device didn't know what happened in the lab, and the lab didn't know what was promised to investors. Employees signed strict confidentiality agreements. Staff believed IT was watching their computers, and assistants friended employees on Facebook to see what they posted.
Even the man who was supposed to sign off on results couldn't stay: lab director Adam Rosendorff resigned in late 2014 and later testified he'd been pressured to vouch for tests he had no confidence in, and that the company was more about PR and fundraising than patient care.
The man enforcing all of it was president Sunny Balwani — also Holmes's boyfriend, living with her. The board wasn't told.
THE NUMBERS NOBODY CHECKED. In 2014 investors were told Theranos would bring in more than $100 million. It made a little over $100,000. That isn't a bad forecast; that's a thousandfold gap between the story and the bank account.
Investors were also told the military was using the technology in Afghanistan and on medevac helicopters. It wasn't. It was easy to believe because General James Mattis sat on the board and had pushed for the device to be tested in war zones. If anyone would know, surely he would.
That's the trap. Every new number made the next question more expensive. The board had lent its reputation. The investors had written enormous checks. Walgreens had customers walking in daily. By now, admitting doubt meant admitting you'd been fooled.
And Theranos made doubt cost even more. It hired David Boies, one of the most feared lawyers in America. Former employees who spoke up got legal letters, and some believed they were being followed.
THE GRANDSON. Tyler Shultz joined in 2013 out of Stanford; he'd met Holmes through his grandfather, George Shultz. In the lab he watched quality checks fail and get rerun until they passed, and bad results thrown out as outliers. In April 2014 he wrote to Holmes laying it out.
He got an answer from Balwani instead — long, cold, condescending — and the only reason anyone replied was his last name. He wrote back: consider this my two weeks notice.
Then he told his grandfather that the machine didn't work and real patients might be getting wrong results. George Shultz — one of the most respected men in American public life — listened to his grandson, and believed Elizabeth Holmes.
One evening Tyler walked into his grandfather's house and found the company's lawyers waiting with papers to sign. He refused. His family spent more than $400,000 defending him, and for well over a year he and his grandfather barely spoke. When George Shultz turned 95 and the family gathered, Tyler wasn't there. Elizabeth Holmes was.
THE ONE WHO WROTE IT DOWN. Erika Cheung was 22, just out of Berkeley, in the same lab. She watched the same quality controls fail, stayed until 3 a.m. trying to make them pass, and kept hearing the same instruction: just get the results out. After about seven months she quit and wrote to federal lab regulators.
Both of them had been talking to the same reporter, John Carreyrou at the Wall Street Journal. When Holmes learned what he was working on, she went over his head to the paper's owner — Rupert Murdoch, her own investor, $125 million in. She later admitted it on the stand. Murdoch told her he trusted his editors.
THE END. October 15, 2015, the story ran. That same day Holmes went on CNBC: "First they think you're crazy, then they fight you, and then all of a sudden you change the world."
For a few weeks plenty of people believed her — of course they did, they'd been believing her for years. But regulators were already moving. Inspectors warned the lab's problems put patients in immediate jeopardy. Theranos voided two years of Edison results, and corrected reports went to doctors and patients who had made health decisions on wrong numbers. Walgreens shut the centers and sued. Regulators banned Holmes from running a lab. Forbes revised her net worth from $4.5 billion to zero.
In 2018 the SEC charged her with what it called a massive fraud. By September, Theranos was gone. In January 2022 a jury found her guilty of defrauding investors; she was sentenced to 11 years and 3 months. Balwani got nearly 13. She is expected to be released in 2030.
SPLASH'S READ: The temptation is to file this under one woman who fooled the world. One person can't hold up a $9 billion lie for twelve years. That takes a crowd.
What the story really shows is a chain of people each pointing at the next: customers trusted Walgreens, Walgreens trusted the money, the money trusted the board, and the board trusted Elizabeth. Nobody was the check. Everybody assumed somebody else had been.
The Forbes cover did more damage than any forged document. Once the valuation was public, doubt became expensive — for the board that had lent its name, for Kering-sized investors who had written nine figures, for the pharmacy chain that couldn't risk CVS getting there first. The lie stopped needing Holmes to defend it. It was being defended by everyone's balance sheet.
And the detail worth holding is the one about family. George Shultz chose the founder over his own grandson, at his own birthday party. That's the real product Theranos sold: not a blood test, but a story powerful enough that smart people would pay — with money, with reputation, with their own blood — to be part of it.