VC-backed startups commit more fraud
Maybe the fundraising system is part of the problem
If startup fraud had a playbook, researchers may have just found it.
A study from Imperial College and Emlyon Business School found out that instead of one big decision, fraud usually grows through a series of tiny decisions, and it’s a 3-stage process called “façading”:
Stage 1 (Surface Façading): This is where founders slightly exaggerate reality, such as rounding up growth/revenue numbers. Roy Lee, the CEO of Cluely, is a good example as in 2025, he admitted on X that the $7 million ARR figure he told TechCrunch wasn’t true.
Stage 2 (Reinforced Façading): Once people ask for proof, founders create fake evidence to support their claim, such as fake contracts, fake customer lists, or fake revenue. This’s exactly what happened at Frank when founder Charlie Javice hired someone to generate millions of fake customer records so that JPMorgan would believe the numbers before buying the company for $175 million.
Stage 3 (Deep Façading): Eventually, the product itself has to maintain the illusion. Nate, Inc. is a recent example as founder Albert Saniger raised more than $42 million while claiming his shopping app was powered by AI, when much of the work was actually being done manually by contract workers.
This is an important finding because they suggest the higher fraud rate is not only about ambitious founders or a few dishonest people, but it seems like once billions of dollars, aggressive growth targets, and investor expectations enter the picture, founders start playing a very different game.
Subscribe, become a member, or gift a subscription to a friend.
The fundraising system may be part of the problem
Indeed, another study from the University of Toronto looked at 654 fraud cases involving U.S. VC-backed startups and found that VC-backed startups are more likely to face fraud charges than bootstrapped startups.
Here are some of the findings:
Startups founded during overheated funding markets, where investors perform weaker due diligence, are 19% more likely to commit fraud later.
Investor’s unrealistic growth expectations create the pressure that pushes founders toward bad decisions.
Startups with founder-controlled boards were about twice as likely to commit fraud.
Founders accused of fraud often have little trouble raising money again in the future.
None of these findings prove that investors cause fraud, but together, they suggest that fraud may not be only about dishonest founders, but it may also be influenced by the way the fundraising system selects founders, expects them, rewards them, and oversees them after the investment.
That completely changed how I thought about startup fraud.
VC may be creating its own version of Gresham’s Law
Gresham’s Law is an old idea from economics. It says that when good money and bad money are treated as if they have the same value, people spend the bad money and keep the good money. Over time, the bad money becomes more common because there is little reason to use the good money.
I think something similar may be happening in startup fundraising.
Imagine 2 founders raising money:
Founder 1 tells investors the honest version of the story. They explain that the company grew 40% last quarter, but they are not sure whether that growth will continue.
Founder 2 confidently claims that the company is about to grow 300%, even though there is little evidence to support it.
During a pitch meeting, both founders are often judged in the same short amount of time. Investors usually cannot verify every claim before making a decision. That means confidence and truth can sometimes look exactly the same.
The honest founder may actually look less impressive because they admit uncertainty, while the founder making bigger claims appears more ambitious and more convincing. If this happens often enough across many fundraising rounds, founders who exaggerate may end up raising more money than founders who tell the complete truth.
That does not mean investors want dishonest founders. It simply means the fundraising process is not always good at telling the difference between genuine confidence and made-up confidence.
Seen through that lens, if honesty isn’t something the market measures very well in the first place, then it makes sense that fraud allegations don’t always stop founders from raising money again.
And if this idea is correct, then the biggest losers are not only the ones who get caught for fraud, but they are also the honest founders who lose funding rounds because someone else was willing to tell a bigger story.
AI has made the hardest part of fraud much easier
Tim Weiss, one of the researchers, pointed out that today’s AI funding boom is the kind of market where this fraud becomes especially dangerous.
Why?
In the past, serious fraud usually required a lot of work.
Charlie Javice reportedly needed people to help create a fake customer database.
Albert Saniger built and managed large teams of workers to make his product appear more automated than it really was.
Trevor Milton at Nikola famously had to move a truck to the top of a hill so it could appear to drive on its own.
Even dishonest founders faced real costs as creating convincing fraud required money, people, and time.
Today, those costs have fallen dramatically.
A founder can now use AI to generate convincing demo videos, create realistic analytics dashboards, write believable customer testimonials, or simulate product interactions in just a few minutes. Many of the things that once required entire teams can now be produced with a handful of prompts.
There is an irony here. Nate reportedly used humans to pretend it had AI. The next generation of fraud may use AI to pretend there is even better AI. It could become cheaper, faster, and much harder to detect.
The highest level of the façading ladder used to be difficult because it required constant effort to maintain the illusion, but it seems AI has removed much of that effort.
So... what do we do with this?
I don’t think the answer is simply telling founders to “be more honest” because that assumes the problem is only about people’s character.
I think the bigger problem is that it’s becoming much easier to fake things, while the way we check whether those things are real hasn’t kept up.
If you're a founder:
You shouldn’t assume losing a fundraising round is proof that your story wasn’t good enough. Sometimes the market rewards confidence more than accuracy. That shouldn’t convince you to start exaggerating your numbers or making promises you can’t support.
Instead, make your proof difficult to fake. Show investors your product working live instead of relying on polished demo videos. Introduce them to real customers instead of sending screenshots of testimonials. In a world where AI can generate almost anything, reality itself becomes a competitive advantage.
If you are evaluating a startup:
Whether you’re an investor, customer, or partner, careful due diligence is important. Committing fraud no longer requires significant time, money, and operational effort because AI has dramatically lowered those costs. That means demos, dashboards, customer quotes, and even product interactions deserve much more scrutiny than they did only a few years ago.
Better founders or better performers?
We spend a lot of time talking about fraud as if it’s a problem of character. The research suggests that systems matter, incentives matter, and the people sitting around the table matter.
Maybe the question isn’t whether founders are becoming more dishonest, but it’s whether our fundraising system has become better at selecting great performers than great builders
If investors cannot reliably tell the difference between evidence and presentation, and AI keeps making presentation cheaper to manufacture, then the qualities that get rewarded slowly drift away from the qualities that actually build enduring companies.
That feels like a much bigger risk than any individual fraud case.
Borrow our audience?