Tyler Shultz Shocks: AI Boom Is 'A Great Opportunity To Go Deceive People'

Tyler Shultz knows what a ten-billion-dollar lie looks like from the inside. The young employee who helped expose Theranos has spent more than a decade watching charisma outrun evidence — and he believes the AI boom is producing a new class of the same con.

In an interview with The San Francisco Standard, Tyler Shultz warned that peak AI mania makes this moment “a great opportunity to go deceive people.” This article unpacks that warning: what Theranos actually got away with, why artificial intelligence is even harder to audit than a blood-testing machine, and what investors, builders, and bystanders should do before the next round closes.

Table of contents

Why Tyler Shultz Is Still Warning People
The Rise and Collapse of Theranos
Why Tyler Shultz Says AI Is Harder to Audit
Trillion-Dollar Valuations, Coffee-Chat Checks
Elizabeth Holmes and Her Second Act
What Tyler Shultz Wants Investors to Remember
A Practical Due Diligence Checklist for AI
A Documentary Moment With an Unsettling Ending
References

Why Tyler Shultz Is Still Warning People

Who is Tyler Shultz?

Tyler Shultz took his first job out of college at Theranos, a role his family made easy to land: his grandfather, former U.S. Secretary of State and Treasury Secretary George Shultz, was one of the company’s most prominent investors and a member of its board. What the young hire saw inside — not a revolution in blood testing but a house of cards — turned him into one of its key whistleblowers.

The interview behind the warning

Tyler Shultz’s latest warning came on an episode of The San Francisco Standard’s podcast, timed to the runaway success of “You Can See Everything,” the Nathan Fielder and Lance Oppenheim documentary capturing Elizabeth Holmes’s final weeks of freedom before prison. His read on the Holmes revival is blunt: “the only reason we’re talking about Elizabeth Holmes right now is because she wants us to be talking about Elizabeth Holmes right now.”

Why the timing matters

The documentary has become one of the highest-grossing documentaries of all time, and it lands at peak AI spending, where companies that have never turned a profit are raising capital at trillion-dollar valuations. For readers already absorbing warnings like the three big types of AI risk experts keep flagging, Tyler Shultz offers something rarer than a prediction: field experience with how tech fraud actually behaves.

What he does now

Since Theranos collapsed, Tyler Shultz has made it his business to warn others — speaking, advising, and pushing for real verification in emerging technology. When he says the AI boom scares him more than Theranos ever did, he is comparing one fraud he watched closely against an entire market he cannot see inside.

What the warning actually says

Tyler Shultz’s argument fits in three sentences, and each one stings. Funding culture has normalized trust without proof. AI makes verification harder than almost any technology before it. And the incentives to fake results have never been larger. If the message sounds familiar to anyone who watched Theranos, that is precisely the point: the same failure modes are back, with more money and fewer guardrails.

The Rise and Collapse of Theranos

A $10 billion promise that never worked

Theranos reached a peak valuation of around $10 billion despite its technology never actually working. Its much-touted proprietary machines performed only a fraction of tests; most samples were quietly processed on commercial third-party equipment while investors were sold a revolution.

Real patients, real harm

This was not a victimless vaporware story. Theranos devices were placed inside Walgreens stores, where faulty results — false positives, false negatives, missed conditions — reached real patients with the authority of a lab report. When Tyler Shultz raised alarms internally and helped journalists confirm the problems, the company answered with surveillance and legal threats.

How the facade fell

The facade cracked in 2015, when journalists armed with internal evidence from Tyler Shultz and other critics began publishing that Theranos was faking results. Prosecutors later convicted Elizabeth Holmes of fraud, and she was sentenced to more than 11 years in federal prison in November 2022.

What makes the timeline grim is that most of Theranos’s backers were not fools. They were wealthy, well-advised people who chose the story over the evidence — the same choice Tyler Shultz now watches investors make again, with far bigger numbers attached.

What investors were told, and what they believed

Backers were repeatedly shown near-magic devices, heard tales of imminent partnerships with big retailers and defense contractors, and asked almost no hard questions about the underlying science. Theranos did not share scientific information or evidence that its machines worked with potential investors, citing trade secrets — and the secrecy filtered out the cautious funds while concentrating the money among the already-converted.

Remember the timeline before scaling any comparison:

How long Theranos ran

Operating before collapse 15 years

Elizabeth Holmes’ prison sentence 11+ years

Faulty Walgreens kiosks live 2 years

Theranos also ran a playbook — and the AI industry runs it back almost move for move:

Playbook move	Theranos, 2003–2018	AI boom echoes
Secretive technology	Proprietary box, zero data shared	“Secret sauce” models behind NDAs
Star-struck backers	DeVos, Walton, political royalty	Celebrity founders, famous angels
Retail enthusiasm	Patients queued at Walgreens	Consumers subscribe to AI monthly
Press kept at arm’s length	Reporters threatened, sources hunted	Skeptics dismissed as “AI doomers”
The verification path	Independent labs exposed it fast	Independent evals exist — demand them

Why Tyler Shultz Says AI Is Harder to Audit

The ChatGPT wrapper problem

“It’s really hard to tell the difference between what is something truly unique and new versus, like, a ChatGPT wrapper,” Tyler Shultz said. And that is precisely the opening he warns about: “If you want to deceive people, I think this is a great opportunity to go deceive people.” A wrapper built on someone else’s model can demo brilliantly while adding nothing defensible underneath.

Trade secrets as a black box

Theranos refused to share scientific information or evidence that its machines worked, citing trade secrets — a shield that warded off many of Silicon Valley’s top venture firms while seducing everyone else. Today’s AI startups run the same play: proprietary architecture, non-disclosure agreements, curated demos, and no access for independent experts. Tyler Shultz’s rule of thumb: secrecy should raise your pulse, not your trust.

The expertise bottleneck

Verifying a blood analyzer takes laboratory scientists; verifying a foundation model takes scarce machine-learning researchers with access to the data, the weights, or at least rigorous third-party evaluations. Most investors have none of that. New startups keep raising on narrative — like Dream, the AI world-models startup that pulled $17M from a16z — while even experts struggle to separate genuine breakthroughs from borrowed plumbing.

Tyler Shultz’s point is that the asymmetry is structural: a fake blood analyzer fails a lab test within weeks, but a fake AI moat fails quietly in production months later — or never visibly at all, because the wrapper keeps riding the model it copied.

Why this is not just a startup problem

The danger reaches far past venture circles. Patients sign up for AI health coaches. Boards approve AI rollouts on vendor promises. Students pay for AI career services. Tyler Shultz spent years watching a fraud that harmed real people one faulty test at a time, and he sees the same asymmetry spreading: enormous trust placed in systems almost nobody can inspect.

Trillion-Dollar Valuations, Coffee-Chat Checks

“$100,000 over a thirty-minute coffee”

“I think the days of having a thirty-minute coffee with someone and sending them $100,000 should be over. Elizabeth took that model and scaled it to $100 million,” Tyler Shultz said. Theranos’ money largely came from individuals — the DeVos and Walton families among them — not the rigorous venture firms that had mostly walked away when shown nothing to verify.

From angel checks to mega-rounds

The scale has since gone cosmic. OpenAI is reportedly working on a $30 billion raise at a $1.25 trillion valuation ahead of an IPO — another company that has never turned a consistent profit commanding nation-sized sums. Tyler Shultz’s fear is that the coffee-chat dynamic now operates at mega-round size, with fear of missing out doing the due diligence.

What hype does to diligence

Commentators keep asking whether today’s AI infrastructure spending is brilliant or a busted bubble. Tyler Shultz answers from the Theranos side: when belief is this cheap, capital stops asking questions — and the founders who stay honest get out-raised by the ones who don’t.

Here is how far one con scaled inside a single company — each step multiplies by ten:

The Theranos money ladder (order of magnitude)

Coffee-chat angel checks $100,000

Holmes’ scaled-up angel model $100 million

Theranos peak valuation $10 billion

Shultz’s suggested shift is unglamorous: fewer dream meetings, more evidence packages; less founder mythology, more expert review. The coffee chat is fine, he argues — the $100,000 wired on the strength of it is not.

Why founders should welcome the scrutiny

Honest builders have everything to gain from the discipline Shultz describes: verifiable claims separate them from the wrapper merchants and let good companies price on substance. If a startup cannot survive a benchmark run by strangers, that is not an unlucky audit — it is the market working exactly as Theranos’s victims deserved and never got.

Elizabeth Holmes and Her Second Act

Haemanthus: “Theranos 2.0”?

Holmes and her partner, hotel heir Billy Evans, are running Haemanthus, a diagnostics company that observers call “Theranos 2.0” and that has reportedly raised about $20 million. Same industry, same origin story, a fresh cap table — and an audience newly primed by a sympathetic film to see her side.

The sympathy strategy

“You Can See Everything” deliberately refuses to characterize Holmes as a master manipulator or a victim, instead showing her uncanny, almost hypnotic knack for making people pity her — even people who know her crimes. Tyler Shultz notes that inviting a camera crew into your final weeks as a mother of two before an eleven-year prison stretch is itself a calculation: pity is more useful for a comeback than hate or mockery.

A second act watched uneasily

“I have this feeling of unease about what might come next,” Tyler Shultz said. “She’s certainly preparing and beginning her second act, and I don’t know exactly where it’s going to go.” In the cultural zeitgeist, Holmes remains more of a joke than a persona non grata — which is precisely why Tyler Shultz keeps saying his warnings out loud.

Tyler Shultz grants the appeal without surrendering to it: Holmes is charismatic, photogenic, and following a script America happily writes for fallen founders — apologize, rebrand, return. His warning is that AI-era media will run that script again, this time around a product even harder to verify.

Why a comeback was even possible

Theranos dissolved, but the affection for its mythology survived. Board members kept their reputations, investors wrote down losses and moved on, and the media cycle kept recycling Holmes as a tragicomic figure rather than a convicted felon. Shultz counts the years between sentencing and comeback and sees an audience eager to forgive a charismatic founder because the alternative — admitting they were fooled — is far less comfortable.

What Tyler Shultz Wants Investors to Remember

Money in someone’s hands is power

“I believe that what Holmes did was raise the biggest angel round of funding of all time,” Tyler Shultz said, “but even at the smaller levels, I think investors need to recognize that when they put that type of money in someone’s hands, they are giving them tremendous power.”

The individual-money problem

Theranos was funded by wealthy individuals rather than venture firms; the funds that demanded scientific evidence never wrote checks. Today, retail investment platforms, social hype, and celebrity capital play the same role for AI startups. Tyler Shultz worries the 2015 lesson — that founders handed unlimited trust will reshape reality to match their pitch — is already being un-learned.

The harassment lesson

When Tyler Shultz tried to sound the alarm, Theranos answered with targeted harassment so intense that he slept with a knife. Companies that attack skeptics instead of answering them tell you everything you need to know about their culture.

He also frames due diligence as self-defense, not hostility: asking a founder for proof does not accuse them of fraud, Tyler Shultz argues. Refusing to ask is what insulted the patients Theranos endangered.

Teach the pattern, not the paranoia

The goal, Shultz stresses, is not to treat every AI founder as a suspect. Most are honest builders racing the same brutal clock. But no one can spot saints or schemers at first glance, so the only honest defense is the evidence habit that banks invented centuries ago and that Theranos investors skipped: trust, then verify — in that order, with the verify part mandatory.

The red-flag list transfers almost one-to-one:

Red flag	Theranos then	Ask this now
No third-party validation	Only internal demonstrations	“Which independent lab tested it?”
Famous board, no experts	Politicians, generals, bankers	“Who can actually read the science?”
Individuals over funds	DeVos and Walton checks	“Which technical investors passed?”
Information lockdown	Trade secrets cited to investors	“Will experts see raw data?”
Attacks on skeptics	Whistleblowers surveilled and sued	“How do ex-employees talk?”

A Practical Due Diligence Checklist for AI

Verify the technology, not the theater

Demos are marketing. Insist on independent benchmarks run by people who did not build the product, customer references you can actually call, and side-by-side comparisons against open models or off-the-shelf APIs — the checks that would have exposed Theranos in a week, had investors bothered to run them.

Follow the money and the board

Ask who else did real diligence and what they found. A board of famous names without relevant expertise — politicians on Theranos’ panel instead of laboratory scientists — is a warning label, and so is a cap table full of individuals who invested on story rather than substance. Tyler Shultz’s own family learned this lesson from inside the board itself.

Protect the messengers

Decide before you invest how a company treats dissent. Internal ethics channels, anonymous reporting, and graceful departures correlate with honesty; non-disclosure agreements weaponized against leavers correlate with what Tyler Shultz endured.

None of this requires personal AI expertise — only the insistence on hiring experts and reading what they report. Public benchmarks, reproducible evaluations, former employees, and regulatory filings all exist now, which is exactly why claims that evade every check deserve the same suspicion Theranos eventually earned.

Make the checklist boring and repeatable

The magic of a checklist is that it does not care how exciting the pitch deck was. Run the same six checks on every deal, write the answers down, and compare them across companies. The pattern that emerges across a portfolio tells you more about a market than the most charismatic single founder ever will.

Six checks every AI investment deserves:

Check	Green signal	Red flag
Independent benchmark	Third-party results replicate	Only curated demos
Customer evidence	Named clients, real outcomes	Anonymous pilots forever
Data provenance	Documented, auditable datasets	“Proprietary data,” no receipts
Board expertise	Technical directors push back	Name-only advisors
Whistleblower posture	Protected internal channels	NDA threats against leavers
Fundraising logic	Milestones drive rounds	Valuation drives the roadmap

A Documentary Moment With an Unsettling Ending

A film that refuses to judge

Directed by Nathan Fielder and Lance Oppenheim, “You Can See Everything” has become one of the highest-grossing documentaries ever precisely because it declines to tell viewers how to feel about Holmes — a narrative choice that lets her charm do what it has always done.

The crimes the film softens

The documentary glosses over the more heinous crimes of Theranos — handing dangerously faulty medical information to real patients, and the targeted harassment of Tyler Shultz when he tried to stop it. As culture re-watches Holmes, she lands more as a joke than a scandal, and few remember the whistleblower who paid for speaking up.

The warning that should stick

AI hype is not unique to AI, and the fix is not pessimism but verification — the same practices that exposed Theranos: angry employees, curious journalists, patient laboratories. Tyler Shultz’s closing unease is worth carrying into every pitch deck you see, because the second act is coming, and this time the product is more powerful than anything the blood-test con artists imagined.

If one sentence survives this cycle, Tyler Shultz says, let it be this: wonder is not evidence. Theranos had wonder; the AI boom has wonder. Every investor, journalist, and employee should demand the boring, checkable proof that the wonder is backed by results.

What to tell your board on Monday

Strip the caution down to two questions you can ask in any meeting, about any vendor or startup: who has verified these claims, and what happened to the last person who doubted them. If either answer is vague, defensive, or hostile — walk. The AI boom will still be here tomorrow, and so will the next whistleblower.