Khosla cut him off with a “talk to the hand” motion and turned to the audience with a wide, this guy amirite? grin. “Here’s a journalist,” he said, “who doesn’t know what’s going on, has an opinion, just like he does, to make interesting stories.” He turned back to Shieber: “I know a lot more about how they’re doing, excuse me, than you do.”

This was in September 2015. And what Khosla didn’t know was that Hampton Creek’s employees and contractors had been covertly buying its jars of eggless mayo from grocery stores for more than a year, allegedly as a way to make the product appear more popular to its retail partners. It would be another year before Bloomberg Businessweek revealed the scheme, in an article featuring an animated GIF of founder Josh Tetrick’s face covered in squirts of mayonnaise. (Hampton Creek has denied wrongdoing, describing the buybacks as quality-control testing. Khosla declined to comment.)

RULE BREAKER: Parker Conrad’s startup Zenefits admitted violating state insurance rules. He’s now starting a new, similar company.Photo: Jim Willson—The New York Times/Redux

The startup community has a set response to this kind of news, and it sounds a lot like Khosla’s sniping. Blindly defend; it’s us against them. After the Wall Street Journal first exposed problems at blood-testing startup Theranos in 2015, for example, venture investors like Greylock’s Josh Elman and Y Combinator’s Sam Altman tweeted defenses against the one-sided “slam piece.”

But as scandals have piled up—and other negative stories have proved to be true—the defensive strategy hasn’t aged well. While some investors are standing by their tainted companies, others are taking pains to distance the bad actors from the rest of the startup pack. Theranos, which has since voided two years of its test results and faces a criminal investigation, is now described as an exception. Just one bad apple. (“Theranos doesn’t represent us, we are better,” a group of startup founders sang in the annual holiday video created by VC firm First Round Capital.) Likewise Zenefits, the human resources startup that admitted its employees had cheated on mandatory compliance training: a freak occurrence. #NotAllStartups.

Lending Club’s loan doctoring? That’s not what startups are about. Same for WrkRiot, the startup that abruptly shut down after an employee accused it of forging wire-transfer documents. Or Skully, the failed maker of smart motorcycle helmets, being sued for “fraudulent bookkeeping.” Or ScoreBig, the struggling ticketing site being sued by brokers. Or Rothenberg Ventures, the firm under investigation after using investors’ money to finance founder Mike Rothenberg’s side startup. (The firm says it informed investors.) Or Faraday Future and Hyperloop One, ambitious, well-funded companies now tainted by lawsuits and accusations of, respectively, overhype and of mismanagement. (Faraday has not commented on its suits; Hyperloop denies the accusation and had settled its suit.) Or any of the dozens of smaller shady accounting shortcuts, growth hacks gone awry, and other implosions too minor to make headlines.

No industry is immune to fraud, and the hotter the business, the more hucksters flock to it. But Silicon Valley has always seen itself as the virtuous outlier, a place where altruistic nerds tolerate capitalism in order to make the world a better place. Suddenly the Valley looks as crooked and greedy as the rest of the business world. And the growing roster of scandal-tainted startups share a theme. Faking it, from marketing exaggerations to outright fraud, feels more prevalent than ever—so much so that it’s time to ask whether startup culture itself is becoming a problem.

“We hope that founders bend the rules but don’t break them. There’s a fine line between entrepreneurship and criminality.”

Fraud is not new in tech, of course. Longtime investors remember when MiniScribe shipped actual bricks inside its hard-disk boxes in an inventory accounting scam in the 1980s. The ’90s and early aughts brought WorldCom, Enron, and the dot-bombs. But today more money is sloshing around ($73 billion in venture capital invested in U.S. startups in 2016, compared with $45 billion at the peak of the dotcom boom, according to PitchBook), there’s less transparency as companies stay private longer (174 private companies are each worth $1 billion or more), and there’s an endless supply of legal gray areas to exploit as technology invades every sector, from fintech and med-tech to auto-tech and ed-tech.

The drama has some investors predicting more disasters. “What if Theranos is the canary in the coal mine?” says Roger McNamee, a 40-year VC veteran and managing director at Elevation Partners. “Everyone is looking at Theranos as an outlier. We may discover it’s not an outlier at all.” That would be bad news, because without trust, the tech industry’s intertwined ecosystem of money, products, and people can’t function. Investors may find the full version of the old proverb is more accurate: “One bad apple spoils the whole barrel.”

Breaking the rules makes you a Silicon Valley hero. That’s great if you’re breaking a dumb rule, not so much if you’re breaking an important one. Startup mythology is packed with stories of That Time Steve Jobs the Genius Did Whatever It Took to Win, and That Time in the 1990s that Larry Ellison the Badass Calculated Revenue the Way He Damn Well Pleased. Today’s founders cite Airbnb’s famous “farming” strategy (it spammed people advertising rentals on Craigslist to lure them to Airbnb). They speak breathlessly about how “T.K.”—Uber cofounder Travis Kalanick—has repeatedly ignored legal roadblocks. Admirers see an aggressive attitude and a $70 billion valuation, ignoring Uber’s careful, behind-the-scenes negotiations with regulators in many cities, notes Bradley Tusk, a political consultant for Uber.

The romantic lone-cowboy tales make it easy for founders to rationalize questionable decisions. “The whole ‘fake it till you make it,’ ‘move fast and break things’ attitude—all those sorts of battle cries are misinterpreted by some folks into making things up,” says Jakub Kostecki, founder of StartupFactCheck, a consultancy that helps investors conduct due diligence on startups. Three-quarters of the 150 early-stage startups he has investigated have pitched investors with misleading or purposely incomplete information, like identifying as “customers” people who are merely using a free trial, or taking full credit for past projects they played only a small role in.

Even when truth-stretching founders get caught, early-stage investors may look the other way. Dave McClure, founding partner of venture fund and accelerator 500 Startups, says misrepresentations don’t always preclude his firm from investing. “You might even find a correlation between ‘interesting’ behavior and successful entrepreneurship,” he says. A founder who recently pitched 500 Startups claimed he “attended” a college he wasn’t even enrolled in—technically true, since he had snuck into some lectures. Fudging the facts is so common at the early stage, it’s practically expected. “Everyone just assumes that the [investment] amounts involved here are too small, [that] reputation matters, and that all startups exaggerate a bit,” says Naval Ravikant, founder and CEO of AngelList, a platform for early-stage investing. AngelList says it has facilitated 1,100 investments over the past six years with no incidents of fraud.

By definition, entrepreneurship requires promoting the heck out of things that don’t exist yet. Even a founder with a strong moral compass and a heart full of good intentions has to persuade investors, engineers, and customers to believe in a future where their totally made-up idea will be real. “That’s not ‘My cola tastes better than yours.’ That’s ‘Let me explain to you how the world’s going to be.’ ” says Chris Bulger, managing director at Bulger Partners, an investment bank that advises technology companies on acquisitions. “Is that person lying when they turn out to be wrong?”

If a founder’s vision does turn out wrong, investors often have little recourse. Ever since Google’s and Facebook’s founders negotiated dual-class share structures to retain control over their companies, hot startups including Uber, Airbnb, Square, Snap, Palantir, and WeWork have pushed for, and gotten, similar founder-friendly terms. If anything goes wrong, too bad. That includes you, Theranos investors: CEO Elizabeth Holmes’s supershares are worth 100 votes per share.

Some founders grow into talented CEOs. Most don’t. That’s an inevitable by-product of Silicon Valley culture, where everybody fetishizes engineers, designers, and inventors while managers get little respect. “We have an epidemic of bad management,” says Phil Libin, a partner at venture firm General Catalyst. “And that makes [bad] behavior more likely, because people are young, inexperienced, and they haven’t seen the patterns before.”

So inexperienced people are handed giant piles of money and told to flout traditions, break rules, and employ magical thinking. What could possibly go wrong? “We hope that entrepreneurs bend the rules but don’t break them,” McClure says. “You know the saying ‘There’s a fine line between genius and insanity’? There’s probably a fine line between entrepreneurship and criminality.”

At its worst, venture capital culture can push founders across that line. To understand VC incentives, flip everything you know about business on its head. Squishy terms like “traction” and “momentum” are more valuable than functional business models, revenue, and profits. But that’s all part of the fun! Venture is high risk, high reward. Wouldn’t you rather play the lottery than toil away for a boring little paycheck for the rest of your boring little life?

To spread risk around, VCs make dozens of bets in each fund. Only one needs to be a Facebook (FB). So why not push the other companies to set impossibly lofty projections? Why not encourage them to advertise a “$1 trillion market opportunity” and “$100 billion in revenue” in their pitch deck? “Every time I meet my investors, they’re asking me, ‘How can we pour more gas on the fire?’ ” one founder recently explained. In public, investors denounce this habit, calling it the “foie gras effect.”

Once the fire is roaring, nobody wants to put it out. While some startups are transparent with their investors—and some investors demand it— the hottest companies have enough leverage to keep inconvenient numbers under wraps. The rich people buying into Uber’s latest round of funding, for example, got no financial information beyond a set of risk factors, according to reports. Likewise, the media feeds on self-reported scraps of information—dubious “annual recurring revenue” here, a growth percentage (from what base?) there. If that sounds familiar, recall the 2000s housing bubble, when Americans reported their incomes on mortgage applications with no outside verification. No surprise, they took liberties. Startup financial disclosures are the “liar loans” of corporate accounting.

It’s easy to shrug off a startup that pushes ethical boundaries a smidgen too far when it’s just a few people and an idea. We assume it will iron things out before it gets big enough to cause real problems. But in the so-called Age of Unicorns, startups can go from zero to $1 billion in valuation in the blink of an eye. And that hype can help them quickly rack up customers, vendors, and employees—all of whom are vulnerable if something melts down. We can’t assume that a billion-dollar valuation is a sign of maturity.

“Startups are desperate,” says Sean Ellis, CEO of collaboration software startup GrowthHackers. “[Mature] companies aren’t going to die if they don’t figure out how to accelerate growth. Most startups will die, and when you’re desperate, you’ll do stupid things.” Like build a computer program to cheat on mandatory compliance training. Or fudge your quarterly numbers. Or buy your own mayonnaise from the store.

Last March, Securities and Exchange Commission chair Mary Jo White traveled to Stanford to deliver a message to Silicon Valley: We’re watching you. The SEC is increasingly concerned, she said, with “eye-popping valuations,” questionable governance, and the lack of transparency at high-risk tech startups.

But when I asked investors about White’s visit, few even remembered it. There’s little reason to worry, the thinking goes, when startups can raise money with ease. Right now the supply of greater fools feels endless. U.S. venture funds are on track to break fundraising records this year, according to PitchBook. Sovereign wealth funds and state-backed investors in the Middle East and Asia are upping their stakes. SoftBank created a $100 billion tech fund alongside Saudi Arabia’s Public Investment Fund. And Fortune 500 companies, betraying their own desperation, are eager to throw money at their disrupters: The number of active corporate venture firms quadrupled between 2012 and 2015, according to CB Insights. The early-stage market is equally flooded. Angel investment in the U.S. grew an estimated 37% between 2009 and 2014, in dollars committed.

Historically, Silicon Valley forgives, even celebrates, failure. E-commerce startup Fab.com promised world domination, then promptly burned through $336 million of investors’ money, selling for just $15 million. That didn’t stop some of the same investors from giving millions to cofounders Jason Goldberg and Bradford Shellhammer for their next startups. (Shellhammer’s failed in less than a year.) Zenefits CEO Parker Conrad stepped down amid the cheating scandal in February; within months he was working on a new employee-benefits startup that sounds a lot like Zenefits. It helps if you spin your meltdown as a learning experience.

But the near-daily revelations of silliness demand greater skepticism toward the truth benders. If America stops trusting the Valley, startups will lose the freedom to innovate. They’ll have a harder time persuading customers, investors, and potential employees to work with them. Their businesses could even be regulated out of existence.

Recklessness with the financial truth is often a sign of an economic bubble about to deflate—see the dot-bombs and Enron in late 2000 and the banks amid the 2007 subprime mortgage crisis. Scandals don’t cause recessions, but they can help trigger one. As White warned her Stanford audience: “Who loses when the truth behind inflated valuations is revealed? I think we all do.”

A version of this article appears in the January 1, 2017 issue of Fortune.