This essay is the result of the author finding himself the subject of a question: how do people closest to a fraud so reliably fail to see it? In other words, “What just happened?” See Author’s Note.
IThe Confession
On the evening of December 10, 2008, the Madoff family gathered in the penthouse apartment at 133 East 64th Street on Manhattan’s Upper East Side. Bernie Madoff, then seventy years old, had asked his sons to come. He had something to tell them.
What happened next has been recounted many times. Mark and Andrew Madoff were not outsiders to their father’s business. They were senior executives at Bernard L. Madoff Investment Securities. They had worked in the firm since the late 1980s. They knew the trading floor, the systems, the personnel. They had spent their entire adult lives inside the enterprise that bore their family name.
And yet, according to their consistent account (and the account that investigators ultimately found no evidence to contradict in criminal proceedings), they had no idea that the investment advisory division of the firm, which operated on a separate floor of the Lipstick Building in midtown Manhattan, was running the largest Ponzi scheme in history.
That evening, their father told them. He said he had been running a fraud. He said it was all a lie. He said there was nothing left.
The sons left the apartment, called a lawyer, and the lawyer called the SEC. The next morning, the FBI arrested Bernie Madoff at his home. When agents told him they were there to determine whether there was an innocent explanation, Madoff replied that there was none.
The world’s reaction was immediate and predictable: How could they not have known?
The question carried an accusation. If they didn’t know, they should have. If they didn’t see, they chose not to look. Anyone too close to the fire must have felt the heat.
That framework is satisfying, and it is often wrong.
It is also freshly relevant. At the time of publication, several new cases have emerged that raise the same questions. These cases are still unfolding, their proceedings in early stages, and this essay does not assess guilt. It examines perception. But the patterns are striking.
In September 2025, Tricolor Holdings, a Dallas-based subprime auto lender that operated roughly sixty-five dealerships, filed for Chapter 7 liquidation after allegations surfaced that the company had been double-pledging collateral, using the same loan portfolios to secure financing from multiple banks simultaneously. Federal prosecutors subsequently charged CEO Daniel Chu and COO David Goodgame with running what the indictment described as a financial crimes enterprise, alleging an approximately $800 million gap between claimed and actual collateral. The company had employed over 1,500 people. The fraud, prosecutors allege, had been operating since at least 2018.
That same month, First Brands Group, a global automotive aftermarket parts supplier employing approximately 26,000 people across five continents, filed for Chapter 11 bankruptcy amid allegations of fabricated invoices, double-pledged receivables, and substantial off- balance-sheet liabilities. Founder and CEO Patrick James resigned within weeks. By January 2026, federal prosecutors had unsealed an indictment charging James and his brother Edward with conspiracy, wire fraud, bank fraud, and money laundering.
In the private credit markets, lenders including BlackRock’s HPS Investment Partners alleged that Bankim Brahmbhatt, the founder of a network of U.S.-based telecom companies operating under the Bankai Group umbrella, had fabricated invoices, forged customer contracts, and created fake email domains to secure more than $500 million in loans. Brahmbhatt’s companies filed for Chapter 11 bankruptcy in August 2025. A federal criminal investigation was subsequently opened. His whereabouts remain disputed.
A separate case involving Radiant World Corporation surfaced in early 2026. The commodities trader is alleged to have fabricated iron ore invoices to secure more than $500 million in financing from the same specialized lending vehicle that had extended credit to First Brands Group. The High Court of England and Wales issued a worldwide freezing order against Radiant World’s assets. The mechanism was identical: the same lender, the same spreadsheet-based invoice fabrication, a different industry. The lender’s own CEO publicly acknowledged the “coincidence of several of these” cases arising in quick succession.
In each of these cases, as in those that came before, people inside the organization who had no hand in the alleged fraud now face the familiar question: How did I not see this? As these matters proceed through restructuring and criminal prosecution, it is an opportune time to examine the cognitive and structural forces that have, time and again, allowed fraud to flourish in plain sight of the people best positioned to catch it.
In every major business fraud of the last several decades, a pattern repeats. A small circle engineers the fraud. A much larger circle, including many personally and professionally intimate with the fraudster, is excluded from it. When the fraud is revealed, the excluded insiders face a devastating double blow: the loss of everything they believed about their professional life, followed by the refusal of many to believe they didn’t know.
This essay is about that double blow and the mechanism that produces it: a blind spot so consistent across cases that it operates with mechanical reliability whenever a charismatic, intelligent person decides to commit fraud within an organization that trusts them.
Understanding the blind spot does not require us to exonerate anyone. What it requires is that we take seriously a possibility that our instincts resist: that the people closest to the deception were not its collaborators but its most perfectly engineered victims.
IIThe Proximity Assumption
Imagine you work for a company. You’ve been there for fifteen years. You know the CEO personally. You understand the business deeply. You’ve been involved in strategic decisions, seen the financial reports, attended the board meetings.
Now someone outside the company suggests your CEO is engaged in massive fraud: a journalist, a short seller, an anonymous tip on a message board. What is your first reaction?
If you are honest with yourself, it is dismissal. The reason is what I will call the proximity assumption: the deeply held, rarely articulated belief that if something were seriously wrong at this company, you, of all people, would know about it. This belief feels reasonable. It is, in fact, one of the most dangerous ideas a person in a position of trust can hold.
But the danger goes beyond the absence of red flags. You have evidence, concrete and tangible, that the allegation is false. Payroll cleared last Friday. The quarterly financials were delivered on schedule. The auditors signed off. The vendor invoices were paid. The client contracts renewed. You did not merely fail to see trouble. You saw, with your own eyes, evidence that there was no trouble.
This is the dimension of the blind spot most frequently overlooked. The insider is not operating in an information vacuum but weighing the outsider’s allegation against a mountain of firsthand, documented evidence to the contrary. The sophisticated fraudster does not simply hide the fraud. They manufacture an affirmative counter- narrative: a steady stream of evidence that the enterprise is healthy. Every payroll that clears, every audit that comes back clean, every quarterly report that shows growth is not the absence of a warning sign. It is the presence of a reassurance. The insider is choosing between an abstract allegation from someone who has never set foot in the building and the concrete, daily experience of a company that appears to be working.
The proximity assumption operates as an invisible shield, deflecting suspicion before it can form. You know this person. You’ve seen no evidence of wrongdoing. Therefore, the rumor must be wrong. The reasoning feels airtight because it rests on genuine experience. But the conclusion, that the absence of evidence in your experience means the absence of the thing itself, is a logical error.
Harry Markopolos spent nearly a decade trying to convince the SEC that Madoff’s returns were mathematically impossible. He was a financial analyst at a competing firm. His submissions were detailed, specific, and correct. They were ignored repeatedly. The SEC’s investigators had met Madoff, visited his offices, spoken to his staff. The proximity assumption did its work: this man is a former chairman of NASDAQ, a pillar of the financial community. The outsider’s math must be wrong.
The proximity assumption does not require stupidity. It requires only trust, and trust is not a failing. It is the foundation on which every functional organization is built. It is the first and most foundational force in the blind spot.
IIIPerception Is Not Observation
To understand why insiders miss fraud that outsiders catch, we must grapple with something fundamental about how the brain processes information. The conventional understanding is that we observe the world and then interpret what we’ve observed. First the data; then the analysis. This understanding is wrong.
Neuroscience and cognitive psychology have established that the brain does not passively receive the world. It actively constructs what we perceive, guided by expectations, mental models, and prior beliefs. Perception is not observation followed by interpretation. Perception is interpretation. What we see is shaped, before it reaches conscious awareness, by what we believe we are looking at.
A kindly old lady walks into a bank, chats warmly with the teller, and walks out having robbed him. No one thinks twice. Later, investigators review the footage and notice her purse was oddly rigid, strangely heavy, concealing something that should have drawn attention. But it didn’t, because nobody expects a kindly old lady to rob a bank. The expectation did not merely influence the interpretation. It determined what people saw in the first place. The brain edited the anomaly out before anyone had a chance to notice it.
Imagine the maître d’ of a busy restaurant. The maître d’ is responsible for the front of the house: seating guests, coordinating the wait staff, and keeping service moving. The owner wants that focus kept on the dining room, away from the kitchen. Orders go in, dishes come out, and the restaurant opens and closes day after day without incident. When the maître d’ hears yelling or loud banging from the kitchen, or sees smoke near the service door, the owner assures the maître d’ that it is normal and under control. The signals are real, but the person has been told they fall outside the job, and the daily routine seems to confirm the reassurance.
The kitchen catches fire because basic safety protocols have not been maintained. The maître d’ has heard the noise and seen the smoke; what has been filtered out is their significance. This illustrates how role-bound attention and repeated reassurance can make a danger present in the environment but absent from a person’s perception.
The maître d’ assumes that those closest to the incidents - the owner and kitchen staff - are best placed to assess the risk, and that everyone shares an interest in the restaurant’s safe and orderly operation.
This is precisely what happens to insiders in a fraud. Their mental model of the CEO, the company, and their own role is so deeply established that contradictory information is not weighed and dismissed. It is not perceived. The anomaly in the financial report, the evasive answer to a reasonable question, the meeting you were not invited to: these are the unusually shaped purse. In hindsight, they are glaring. In the moment, they are invisible.
This distinction matters enormously. It separates a moral accusation from a scientific explanation. To accuse insiders of willful blindness is to say they saw the evidence and chose to ignore it. That is a claim about character. The evidence from cognitive science points to something different: they did not see the evidence at all. Their brains, operating as brains are designed to operate, filtered the anomalies out before conscious evaluation began.
Willful blindness is a choice. Perceptual filtering is a mechanism. The two produce identical outcomes but arise from different processes. The fraudster needs only to maintain conditions under which the insider’s brain does the filtering automatically. This distinction between choice and mechanism is the key to understanding why moral condemnation of the excluded insider so often misses the mark.
The psychologist Daniel Simons demonstrated this in his famous “invisible gorilla” experiment, in which subjects counting basketball passes failed to notice a person in a gorilla suit walking through the scene for nine full seconds. Now imagine that experiment running not for sixty seconds but for fifteen years. Imagine that the gorilla is the possibility that the person you trust most in your professional life is a criminal. The brain’s capacity to filter out that possibility is, under these conditions, nearly absolute.
IVThe Architecture of Exclusion
What makes the blind spot so effective is that it is not merely passive. The smartest fraudsters understand it intuitively and build their operations to exploit it. They do so through separation, whether physical, informational, or social. The instrument varies; the principle does not.
Size alone does not create a blind spot, but a lean, tightly controlled company may leave fewer independent eyes on how information moves between functions. When staff are stretched, the people who might question a report are occupied keeping operations running. An operations manager, for example, may have neither the time nor the mandate to scrutinize the finance team’s work. If information is also controlled from the top, this lack of capacity can make compartmentalization easier to maintain: each function sees enough to do its job, but too little to test the full picture.
Consider the physical layout of Madoff’s offices. Bernard L. Madoff Investment Securities occupied three floors of the Lipstick Building at 885 Third Avenue. The legitimate trading operation, where Mark and Andrew worked, occupied the 19th floor. The Ponzi scheme operated on the 17th. The physical separation meant that the sons could spend entire days at the office without encountering the people manufacturing fake trade confirmations. The 17th floor was, in the words of one former employee, a world unto itself: a small, quiet operation staffed by longtime loyalists who answered only to Bernie.
The separation was not just physical. It was informational. The trading division had its own systems, its own compliance processes, its own revenue streams. If you worked on the 19th floor, everything you could see was legitimate. The data was real. The trades were real. The commissions were real. There was nothing in your daily experience to suggest anything was wrong, because within the boundaries of your experience, nothing was.
This is the architecture of exclusion: the fraudster’s most powerful tool. The goal is not to deceive through false information but to supply enough true information that the insider never thinks to ask about what they are not seeing. The most effective lie is an omission, surrounded by truth.
Elizabeth Holmes perfected a different version at Theranos, separating the fraud informationally by forbidding employees from communicating across departmental lines. In the fall of 2015, a young scientist named Erika Cheung was running quality control checks on the company’s proprietary blood-testing device. The results were inconsistent, the error rates far outside acceptable clinical parameters. She flagged the problem and documented the discrepancies. She was told, in essence, not to worry about it. The problems she was seeing were local. Other areas were working. The overall system was sound.
What Cheung could not know was that every other team at Theranos was having the same experience. Every team hit walls. Every team was told their specific problem was local while the broader system hummed along. The fiction was maintained not by one big lie told to everyone but by dozens of small lies, each tailored to a specific audience, each just plausible enough to hold. The more talented the employee, the more sophisticated the partial truth had to be.
Enron employed the same principle, though its instrument was complexity rather than secrecy. The labyrinthine structure of special- purpose entities and off-balance-sheet partnerships was so intricate that even sophisticated financial analysts could not untangle it. The complexity served a dual purpose: it hid the fraud from regulators and investors, and from most of Enron’s own employees. The people on the trading floor were dealing in real energy markets, executing real trades, making real money. The rot was in a different department, using financial instruments most employees were not expected to understand. And Enron’s culture of intellectual intimidation ensured that anyone who admitted confusion about the company’s more exotic financial structures risked being seen as not smart enough to work there.
Sherron Watkins, the Enron vice president who warned CEO Ken Lay about accounting irregularities, experienced this firsthand. She had the credentials, the access, and the intelligence to see that something was wrong. She raised the alarm and was dismissed, not because her concerns were investigated, but because the culture had established that questioning Enron’s methods was itself a sign of failure. Insiders who did see something lacked credibility precisely because they were insiders: if the system were really broken, surely someone more senior would have noticed.
If Madoff, Holmes, and Enron exploited physical and informational separation, FTX demonstrated that social closeness can serve the same exclusionary function. Key employees lived together in a luxury penthouse compound in the Bahamas. Sam Bankman-Fried cultivated an image of radical transparency: the disheveled genius who slept on a beanbag, who wore cargo shorts to meetings with regulators. The message was unmistakable: we have nothing to hide.
But the communal environment functioned as a mechanism of control. When your boss is also your roommate, the social cost of asking hard questions becomes enormous. You would be questioning not just a business decision but a personal relationship, a community. And the critical nexus of the fraud, the relationship between FTX and Alameda Research through which billions in customer funds were misappropriated, was managed by a tiny group within the larger household.
When John Ray III, the restructuring specialist who had previously overseen Enron’s liquidation, took over as CEO of the bankrupt FTX, he said that in his entire career he had never seen such a complete failure of corporate controls. Yet dozens of people had worked at FTX in good faith. Engineers, compliance staff, marketing professionals. They lived in the same compound, attended the same meetings, saw Bankman-Fried every day. And most of them had no idea that the company was systematically looting its customers. The exchange processed billions of dollars in legitimate trades. The affirmative counter-narrative was, as always, constantly replenished.
Wirecard, the German payments company, offers the starkest illustration. The company fabricated the existence of 1.9 billion euros in cash that did not exist. For years, auditors, employees, and German financial regulators accepted the fiction. Wirecard had a legitimate payments processing business that employed thousands and generated real revenue. It also had a separate division, run by COO Jan Marsalek, that existed largely on paper. The employees in the real business had no way to know. The affirmative counter-narrative (the legitimate operations that made the fraud seem implausible) was as strong at Wirecard as at any company in this essay.
The most extraordinary chapter in the Wirecard saga involved BaFin, the German financial regulator. When the Financial Times began publishing investigative reports questioning Wirecard’s accounting, BaFin’s response was not to investigate the company but the journalists. The regulator filed a criminal complaint against the reporters, accusing them of market manipulation. This is the blind spot at an institutional level: BaFin’s mental model was so deeply embedded that it inverted the investigative instinct. The people raising the alarm were treated as the enemy.
VThe Competence Trap
There is a cruel irony embedded in the blind spot: the more competent you are, the more vulnerable you become.
Insiders at these companies were brilliant at their jobs. The traders on Madoff’s 19th floor were skilled professionals. The scientists at Theranos were serious researchers. Enron’s analysts were among the sharpest in the energy business. FTX’s engineers built technology that actually worked.
Their competence created a trap. Because they were good at what they did, and because what they did produced real results, they naturally assumed the larger system must also be sound. If I am doing excellent work, and my colleagues seem to be doing excellent work, and the company is succeeding, then the company must be what it appears to be. The reasoning is intuitive, and in ordinary circumstances correct. The problem is that fraud is precisely the circumstance in which it is not, and the insider’s competence makes them less likely to consider that possibility.
Psychologists call a related phenomenon the illusion of explanatory depth. People consistently overestimate how well they understand complex systems. Asked to explain how a toilet works or a helicopter flies, they rate their understanding as high, then attempt the explanation, and their confidence collapses. We walk through the world feeling we understand things with which we have only a surface acquaintance.
In a corporate setting, this illusion is amplified by the proximity assumption. You think you understand your company because you are there, because you have attended the meetings, read the reports, walked the halls. All of this genuine experience creates a powerful sense of comprehensive understanding that is nothing of the kind. You understand your corner of the business. You understand what you have been shown. The rest is inference, shaped by trust.
Cynthia Cooper’s story at WorldCom provides an illuminating counterpoint. As the company’s vice president of internal audit, she ultimately discovered that CEO Bernie Ebbers and CFO Scott Sullivan had inflated the company’s assets by over eleven billion dollars.
The discovery was not the result of vigilance or superior skepticism. By Cooper’s own account, it was accidental. She and her team stumbled onto irregularities in capital expenditure accounts and followed them, reluctantly, to their source. She was warned off by Sullivan. She was told she was out of her lane. She continued not because she was immune to the blind spot but because her structural position (running an internal audit function technically adversarial to the CFO’s office) gave her just enough institutional distance to keep asking questions.
What made Cooper different was not character, though she had extraordinary courage. It was position. She occupied a role inside the company but at an angle to its leadership. Close enough to find the discrepancies, but not so socially embedded with the C-suite that the perceptual filtering described in Section III suppressed them. The lesson is that the architecture of her role partially protected her from the forces that blinded everyone else.
This distinction is critical. If the blind spot were a matter of moral fiber, the prescription would be easy: hire more honest people. But the blind spot is a product of structural position and perceptual filtering, and the prescription is more demanding. It requires changing the architecture of organizations so that the cognitive conditions for seeing are built into the system, rather than left to the courage of individuals.
VIThe Outsider’s Advantage
There is a striking consistency in who eventually catches major frauds. It is rarely an insider.
Jim Chanos identified Enron’s accounting problems before the company collapsed. John Carreyrou broke the Theranos story. Harry Markopolos spent a decade trying to expose Madoff. Dan McCrum of the Financial Times pursued Wirecard for years. The crypto journalist who published the CoinDesk report on Alameda Research’s balance sheet set off the chain of events that destroyed FTX. These individuals were not smarter than the insiders. What they had was more fundamental: they lacked the proximity assumption. They approached without the accumulated weight of personal experience, social relationships, and identity entanglement. They started from the numbers and asked a simple question: Does this make sense?
As described in Section II, Markopolos’s analysis of Madoff’s returns required no insider information, no forensic sophistication. It required only the willingness to do arithmetic without the distorting lens of trust. That willingness is structural, not moral. It is available to outsiders precisely because they have not spent years constructing the mental model that forecloses the question.
The SEC’s failure with Madoff illustrates the worst of both worlds: enough familiarity to generate trust, not enough access to see past it. The outsider’s advantage is structural. Outsiders lack the mental model that insiders have spent years constructing. They do not know the CEO personally. They have not attended the holiday parties or shared the victories. They approach claims with the skepticism that insiders have been trained, by proximity itself, to suppress. And because their identities and careers are not intertwined with the company’s success, they can afford to see what the data actually shows.
VIIThe Entanglement of Self
The most powerful force in the blind spot is not cognitive but emotional: what the insider stands to lose.
The conventional explanation for why insiders fail to see fraud is financial self-interest. They were getting rich, the reasoning goes, and so they looked the other way. But this explanation does not survive contact with the actual cases. In most major frauds, the excluded insider is not someone whose wealth depends on the fraud continuing. These are people who could be doing just as well at other firms. Their skills are portable. Their reputations are strong.
More to the point, the fraud itself is actively harmful to them. When the scheme collapses, and every scheme eventually collapses, it takes the insider’s career with it. The years of good work are recast as complicity. The professional network becomes radioactive. The references are worthless. The Madoff sons were not beneficiaries of the Ponzi scheme. They were its most devastated casualties. The fraud did not enrich them. It destroyed them.
If the entanglement is not about money, what is it about? Trust, and what the failure of trust means for the insider’s understanding of themselves. When you have worked alongside someone for years, when you have staked your professional judgment on the proposition that this person is worthy of trust, the discovery of fraud does not merely reveal a crime. It reveals a catastrophic failure of your own perception. You sat across from someone at a thousand meetings, evaluated their character, and got it completely wrong.
For most people, the capacity to judge character is central to their identity. To discover that this capacity failed spectacularly, in precisely the domain where you felt most expert, is not a professional setback. It is an existential crisis. It calls into question not just what you knew but who you are.
This is the true entanglement, and it operates with enormous force because it has nothing to do with money. The brain, confronted with information that threatens to demolish the insider’s understanding of their own judgment, intercepts it before it can be fully processed. The mechanism is the perceptual filtering described in Section III, but the motive force is different. The threat is not the fraud itself, which the brain has not yet recognized. The threat is the possibility of the fraud, which would shatter a foundational belief in one’s own powers of perception.
Consider how difficult it is for a person to recognize that a long- term romantic partner has been unfaithful. The evidence, in retrospect, is overwhelming. Friends see it clearly. The betrayed partner, with the most data and the closest vantage point, is the last to know. We understand intuitively that their emotional investment restructured their perception. The same dynamic operates in professional relationships of deep trust. The insider does not look away because looking would cost them money. They do not look because looking would cost them something far more precious: the belief that they are the kind of person who would have seen it coming.
Mark Madoff hanged himself on December 11, 2010, exactly two years after his father’s arrest. Andrew Madoff died of lymphoma in 2014, still fighting publicly to clear his name. Whether the sons knew, whether Ruth knew, whether Peter Madoff (who served as the firm’s chief compliance officer and who did plead guilty) acted with full awareness: these questions may never be resolved with certainty. What is certain is that the cost of proximity was catastrophic, and that the cost had nothing to do with financial dependence.
VIIIThe Social Proof Machine
Individual perception does not operate in isolation. We calibrate our sense of reality against the people around us. If everyone in the room appears calm, we assume there is no reason to panic. If no one is raising concerns, we assume there are none to raise. Psychologists call this social proof, and in corporate fraud it functions as an accelerant, amplifying every other force that produces the blind spot.
At Enron, Fortune named the company “America’s Most Innovative Company” for six consecutive years. Wall Street analysts praised its leadership. Arthur Andersen, one of the most prestigious accounting firms in the world, signed off on its financial statements. If you harbored doubts, you were confronted with a wall of external validation that made those doubts feel irrational.
At Theranos, the social proof was even more extraordinary. The company’s board included Henry Kissinger, George Shultz, James Mattis, and Sam Nunn. Holmes had appeared on the cover of Forbes and been celebrated at conferences alongside the titans of Silicon Valley. If you were a scientist whose device was not performing as promised, the sheer weight of public reputation made it far easier to believe the problem was local, your module, your calibration, your technique, than systemic. The social proof did not just reinforce the blind spot. It made the blind spot feel like clear-eyed realism.
Tyler Shultz, the grandson of board member George Shultz, is one of the few insiders at Theranos who raised the alarm. After he contacted regulators, his own grandfather’s associates pressured him to recant.
The elder Shultz, a former Secretary of State, continued to support Holmes publicly even after his grandson raised concerns. A family member with firsthand knowledge was less credible, within the organization’s social system, than the accumulated reputation of the founder. This is the social proof machine at its most extreme.
The blind spot is not an individual cognitive failure. It is a system in which each layer, the proximity assumption, the architecture of exclusion, identity entanglement, the competence trap, and social proof, strengthens the next, and the composite effect can persist for decades in the presence of intelligent, honest, attentive people. But the system is not only cognitive. It is also institutional.
IXThe Institutional Accelerant
The blind spot operates inside institutions, and those institutions have incentive structures that, even absent malice, systematically reinforce every perceptual distortion described above.
Consider how the typical corporation rewards its employees. Compensation is tied to the company’s reported performance. Promotions go to those who advance the leadership’s agenda. Prestige accrues to the team players, the people who solve problems rather than raise them. Stock options and bonuses create a direct financial interest in the company’s valuation, which means a direct interest in believing the numbers are real. None of this requires fraud or malice. It is the normal operating logic of a competitive enterprise, and it is perfectly designed to punish skepticism and reward credulity.
The employee who raises a concern about the CFO’s numbers is not rewarded for vigilance. They are treated as a problem: moved off the project, excluded from meetings, marked as not a team player. Sherron Watkins at Enron, Erika Cheung at Theranos, and anonymous whistleblowers in countless other cases describe the same experience: the institutional immune system treated the person who identified the disease as the disease itself.
Performance pressure compounds the problem. In an environment where quarterly targets must be met, where covenant compliance must be demonstrated, where investor expectations must be satisfied, the institutional appetite for bad news is near zero. The system does not need to instruct employees to look the other way. It structures their world so that looking the other way leads to bonuses, promotions, and continued employment, while looking too closely leads to friction, isolation, and career risk.
This is not a conspiracy. It is something more insidious: ordinary institutional incentives that, in the presence of fraud, function as co- conspirators. The blind spot is cognitive, but the institution provides the greenhouse in which it grows. And these same incentive structures are, in a fraud-free company, entirely rational. The employee who focuses on their work, trusts their leadership, and does not second-guess every financial report is, in a healthy organization, doing exactly what they should. The institutional accelerant is invisible because it is indistinguishable, in real time, from normal operation.
All of this has so far been described from the perspective of the innocent insider. But the perpetrator is not immune.
XThe Perpetrator’s Blind Spot
The blind spot has a mirror image, one that operates on the perpetrators themselves. It helps explain not only why frauds succeed but why they end the way they do: not in careful retreat but in spectacular, often preventable collapse.
The conventional image of the fraudster is cold-eyed calculation. In the early stages, that image may be accurate. The initial deception requires hyperawareness: constant monitoring of what each person knows, what each document reveals. But fraud is a sustained campaign, and the longer it continues, the more the perpetrator’s own cognition warps under the weight of maintaining it.
What happens is operational myopia. The daily mechanics of fabricating documents, managing money flows, and keeping stories consistent become so all-consuming that the perpetrator loses the ability to assess the broader landscape. Psychologists call this attentional narrowing: under extreme cognitive load, the brain sacrifices peripheral awareness for focused attention on the immediate task. The fraudster, managing an increasingly precarious web of deception, focuses so intently on the next fabrication that they lose awareness of the signals that the structure is about to give way.
Consider Madoff in autumn 2008. The global financial system was in freefall. Hedge funds were reporting catastrophic losses. And Madoff’s fund, which purported to use a strategy deeply exposed to equity markets, reported modest positive returns. In any other year, the consistency was suspicious. In 2008, it was absurd. Yet Madoff did not adjust. He was so embedded in the operational rhythm that he could not see how radically the external context had changed.
The late-stage sloppiness is, case after case, disproportionate to the sophistication of the early stages. Wirecard’s fraud fooled major accounting firms for years, yet by the end Marsalek was relying on forged bank statements a first-year analyst could have flagged. Holmes launched a Walgreens partnership that put unreliable devices in front of real patients, generating the evidentiary foundation of her prosecution. Bankman-Fried publicly sparred with the one person on earth with the most power to destroy FTX. None were calculated risks. They were the actions of people whose field of vision had contracted to the next reassurance, the next day of survival.
There is a deeper mechanism. Over time, the perpetrator becomes a victim of their own counter-narrative. The fraud started as something they knew to be false. But surrounded by believers, rewarded for telling the story, the boundary between fiction and belief begins to erode. They do not forget the fraud exists. They begin to believe it is a temporary expedient the underlying business will justify. The lie becomes, in their mind, a premature truth.
Eugene Soltes, a Harvard Business School professor who spent seven years corresponding with nearly fifty convicted white-collar executives, found empirical support for this account. The critical finding in his 2016 book Why They Do It is that these executives never deeply felt their decisions were harmful. The distance between their actions and their consequences, a distance built into modern corporate life, rendered the harm invisible to intuitive judgment even as their rational minds knew the conduct was wrong.
Donald Cressey, whose interviews with incarcerated embezzlers in the 1950s produced the foundational research on fraud motivation, identified the same phenomenon decades earlier. Each perpetrator developed a prior rationalization that allowed them to reconcile the criminal act with their self-image. They were borrowing, not stealing. Protecting the company, not defrauding it. The rationalization was a lie told to themselves, and they believed it even as the factual predicate eroded.
Fastow, Enron’s chief financial officer, illustrated this in his conversations with Soltes. He described himself as doing what he was incentivized to do, finding ways around the rules through complex processes that exploited loopholes. The rules were obstacles; the fraud was creativity; the accounting fictions were innovations everyone up the chain had endorsed. He knew the structures were deceptive. He believed the deception was sanctioned. The knowledge and the belief coexisted, and the belief won because it was more compatible with the life he wanted to be living.
A person can simultaneously know that what they are doing is wrong and not feel that it is wrong. This is not hypocrisy but a well- documented feature of human cognition. Fraud unfolds not in a single dramatic act but in thousands of individually unremarkable decisions spread across months and years. There is no single moment of reckoning. Each act is small enough to be absorbed by the rationalization. The catastrophe is cumulative, but the experience of it is granular.
But none of this erases the perpetrator’s culpability. Their situation is categorically different from the innocent insider’s. The insider has no underlying knowledge to be distorted. The perpetrator possesses certainty: they know the invoices are fabricated because they fabricated them. What the blind spot erodes is not the knowledge but the will to act on it. The perpetrator chose, at every stage, to continue rather than confess, to double down rather than stop. That these choices were made within a fog of self-deception does not make them less choices.
The result is a bitter irony. The fraudster who began as the clearest-eyed person in the room ends as someone nearly as blind as the people they deceived. They built the blind spot as a tool. In the end, they inhabited it.
XIThe Gray Zone
Between the innocent insider who sees nothing and the architect who designs the fraud, there is a third category: the subordinate who participates, knowingly but under direction, in conduct they understand to be wrong. These are not masterminds. Neither are they unwitting bystanders. They occupy a gray zone that is, in some respects, the most troubling territory in the landscape of fraud, because it is the territory that most of us, under the right circumstances, could imagine inhabiting.
The gray zone is defined by authority, dependence, and a sincere belief that the situation is temporary. The subordinate’s superior controls their livelihood, their professional identity, and often their social world. The instruction comes wrapped in reassurance: this is a one-time adjustment, we will fix it next quarter, no one will be harmed.
The subordinate sees the line. They know they are crossing it. What they cannot see is that the line will never be uncrossed.
The pattern is remarkably consistent across cases. At Enron, Michael Kopper, a managing director who worked directly under CFO Andrew Fastow, became the first cooperating witness. He pleaded guilty in August 2002, admitting to conspiracy and money laundering, and agreed to forfeit $12 million. Ben Glisan, Enron’s treasurer, pleaded guilty the following year and became the first Enron executive to serve prison time. Kopper received three years and one month; Glisan received five years. Both testified against Fastow, who had recruited and directed them.
At FTX, Caroline Ellison, who served as CEO of Alameda Research, pleaded guilty in December 2022 to seven counts including conspiracy and wire fraud. Gary Wang, FTX’s co-founder and chief technology officer, pleaded guilty to four counts; he had written the code that allowed Alameda to withdraw unlimited funds from the exchange. Nishad Singh, the chief engineer, pleaded guilty to six counts, having confronted Bankman-Fried about the missing funds in September 2022. Ellison received two years where Bankman-Fried received twenty-five. Singh, who cooperated extensively, was spared prison entirely.
At First Brands, the vice president of finance and the company’s former CFO have each pleaded guilty to charges including conspiracy, wire fraud, and bank fraud, and are cooperating with prosecutors in the case against the company’s principals. Read together, their pleas and cooperation show the same pattern seen at Enron and FTX: subordinates who admitted participating in the fraud are cooperating with the government against the principals prosecutors say directed it.
The legal system resolves this ambiguity through the plea agreement, which serves as a rough sorting device for moral culpability. The cooperating subordinate admits wrongdoing and provides testimony that allows the government to reach the principals. In exchange, the subordinate receives a sentence reflecting both their participation and their lesser role. The sentencing disparity is the system’s imperfect acknowledgment that not all participants bear the same moral weight.
But the moral question remains open. Nishad Singh told the court he was overwhelmed with remorse and had strayed from his values. Ellison said her brain could not comprehend the scale of the harm. These statements may be strategic. They may also be true. What they describe is the experience of a person who crossed a line they could see, who knew they were crossing it, and who did so because the forces arrayed against compliance, authority, loyalty, fear, and the belief that the situation was temporary, were stronger in the moment than the forces on the side of refusal.
The law is clear that obedience is not a defense. If subordinates could escape liability by pointing upward, the architecture of fraud would be impregnable. But there is a spectrum. At one end is the sales representative who closes deals in good faith, unaware that invoices will later be fabricated. At the other is the CFO who personally directs the fabrication. Between them are people whose knowledge shades from the unwitting to the complicit: the accountant told not to worry about irregularities, the controller preparing reports from numbers they have been instructed not to question, the finance executive who knows the documents are false because they helped produce them at the CEO’s direction.
The gray zone is the most unsettling territory in the landscape of fraud, because it is the territory most of us would occupy if the circumstances aligned. Most of us will never be Bernie Madoff or Elizabeth Holmes. But most of us can imagine, with uncomfortable ease, being the person in the room when the CEO says: I need you to do this. It’s temporary. We’ll fix it. And most of us are not entirely certain what we would do.
XIIThe Digital Accelerant
The cases in this essay span three decades, from Madoff’s Ponzi scheme to the collapses of 2025. Over that period, the tools available to fraudsters and the institutions that might catch them have been transformed by technology. The transformation has not been neutral. On balance, it has made the blind spot wider.
Algorithmic opacity is the most obvious accelerant. When a company’s revenue flows through automated systems, when its financial reporting is generated by software that only a handful of engineers understand, the architecture of exclusion becomes self- executing. The separation that Madoff achieved through physical floors and Holmes through departmental silos can now be achieved through code. An employee who cannot read the algorithm that generates the numbers they report on is in the same position as Erika Cheung at Theranos: surrounded by partial information, unable to see the whole.
Digital fabrication has also lowered the cost and raised the quality of forgery. The crude fake bank statements that exposed Wirecard are becoming relics. AI-assisted document generation, synthetic invoices, deepfaked confirmations, and algorithmically generated customer records are within reach of a determined fraudster with modest technical resources. The affirmative counter-narrative can now be manufactured at scale, with a verisimilitude that makes detection by human review increasingly difficult.
The Radiant World matter, which surfaced within months of the First Brands collapse, shows how readily the same mechanism runs in a new setting. The alleged fraud targeted the same lending vehicle, relied on the same spreadsheet-based invoice fabrication, and produced the same audit failure. Different industry, different perpetrators, same architecture; the same lender on the other side. The digital layer does not invent new frauds. It allows the same fraud to be reproduced at lower cost and, until caught, with greater surface plausibility.
Most importantly, technology has increased the psychological distance between the perpetrator and the harm. When fraud required physically forging a signature or looking a lender in the eye, the act retained a visceral concreteness that made the rationalization harder to sustain. When it requires only a keystroke, when the fabricated invoice is generated by a script and the fictitious customer exists only as a row in a database, the distance that Soltes identified as rendering harm invisible to intuitive judgment grows wider still. The digital layer does not create the blind spot. But it thickens it, for perpetrators and insiders alike, in ways that the governance structures of most organizations have not yet adapted to address.
XIIILimits and Boundaries
A theory that claims too much is worse than no theory at all. This essay has argued that the forces it describes operate with something close to mechanical reliability. But there are boundary conditions.
First, insiders sometimes do see early. Sherron Watkins at Enron, Erika Cheung at Theranos, and Nishad Singh at FTX all perceived that something was wrong while still inside the organization. Their experiences are not counterexamples to the blind spot thesis; they illustrate its limits. In each case, the insider who saw occupied a particular structural position: Watkins had been recently reassigned to a role with a new vantage point; Cheung was running quality control that forced direct contact with failing results; Singh was close enough to the inner circle to observe fund flows. The blind spot is powerful, but it is not omnipotent. Structural disruptions (whether a change in role, a new data point, or an anomaly too large to filter) can puncture it.
Second, proximity sometimes helps detection. Internal auditors like Cynthia Cooper, compliance officers who report outside the chain of command, and finance professionals with access to raw data can occupy the rare sweet spot: close enough to see the evidence, independent enough to evaluate it. The lesson is not that proximity is always blinding, but that it is blinding under specific conditions: when the insider is socially embedded with leadership, when their identity is intertwined with the organization’s success, and when the affirmative counter-narrative is strong enough to outweigh ambient signals.
Third, this essay has focused on cases where fraud was eventually discovered and insiders were genuinely surprised. Sometimes the best evidence that insiders were genuinely surprised is that well-regarded, highly qualified personnel were still at the company when the fraud was uncovered. Survivorship bias applies: there may be cases where insiders quietly noticed and quietly left, or where low-level concerns were resolved before the fraud grew large. The blind spot is a dominant pattern, not a universal law.
XIVSeeing Through the Blind Spot
If the blind spot were a matter of bad character, the solution would be simple: hire better people. If it were a matter of insufficient information, the solution would be transparency. But the blind spot is a structural and cognitive phenomenon that exploits the normal functioning of the human brain. The solutions must be structural and cognitive as well. None eliminate the blind spot. They only narrow it. But in a domain where the difference between seeing and not seeing is measured in billions of dollars, narrowing is enough to justify the effort.
Structural independence of oversight. The first lesson is genuinely independent oversight: not the kind that exists on paper (Enron had an audit committee; Theranos had a board of luminaries; Wirecard had auditors from EY) but oversight exercised by people not socially embedded with leadership, who do not owe their positions to the CEO, and whose incentives are aligned with finding problems. Cynthia Cooper’s internal audit function reported to the board’s audit committee rather than to the CFO whose fraud she uncovered. That structural separation was the difference between the fraud being caught from within and continuing indefinitely.
Separation of verification from operations. The architecture of exclusion works because the people who could verify the numbers depend on those numbers being good. Effective governance separates these functions: independent valuation of assets, third-party confirmation of receivables, segregation of duties in financial reporting, and audit trails accessible to people outside the chain of command. Critical verifications should be performed by parties with no economic interest in the outcome.
Institutional protection for dissent. The incentive structure described in Section IX punishes skepticism and rewards conformity. Reversing this requires more than a whistleblower hotline. It requires visible, consequential protection for those who raise concerns: direct reporting lines to independent board committees, legal protections beyond minimum statutory requirements, and a culture in which questioning the numbers is a professional obligation rather than a career risk. Watkins, Shultz, and Cheung all faced retaliation. The formal mechanisms are insufficient without institutional commitment to their enforcement.
Cultivated outsider perspective. Organizations should deliberately seek out the views of people who lack the proximity assumption: short sellers, investigative journalists, industry skeptics, anonymous whistleblowers. The instinct to dismiss outside critics as uninformed or hostile is the blind spot in action. Effective governance institutionalizes the opposite: requiring that external criticism be formally investigated, not reflexively dismissed, and that results be reported to independent directors.
Rotation and structural disruption. Tenure is a risk factor. The longer an insider occupies a single vantage point, the more entrenched their mental model and the more powerful the perceptual filtering. Mandatory rotation of key oversight roles, periodic engagement of new auditors, and deliberate reassignment across divisions introduce the fresh perspective that punctures settled assumptions.
Technological countermeasures. If technology has made the blind spot wider, technology can also narrow it. Automated anomaly detection, continuous auditing systems, and AI-assisted pattern recognition can flag inconsistencies that human reviewers would miss.
But these tools must be governed by independent parties. An anomaly detection system controlled by the executives it is meant to monitor is no safeguard at all.
Personal cognitive discipline. The most difficult prescription is personal: the regular practice of questioning what we are most certain of. Not in a paranoid way. Not in a way that destroys the trust on which productive work depends. But in a way that acknowledges that our certainty is itself a product of forces operating below conscious awareness. The more confident you are that everything is fine, the more carefully you should examine that confidence.
There is a sentence that echoes through the aftermath of every major fraud. Employees, board members, family members, regulators, investors, all say it with bewilderment and anguish: “I had no idea.”
The conventional response is disbelief. The response this essay proposes is different. It is to take the sentence seriously and to recognize that the mechanisms of human cognition make it not only possible but probable. The people closest to fraud are often the last to see it, not because they are complicit or careless, but because they inhabit a carefully constructed world in which the evidence of their daily experience tells them, affirmatively and repeatedly, that everything is fine.
The blind spot is not a failure of ethics or intelligence. It is a predictable consequence of trust operating inside closed systems, reinforced by institutional incentives, amplified by social proof, and thickened by the digital opacity of modern commerce. That is this essay’s central claim, stated plainly: proximity to fraud does not produce knowledge of fraud. Under identifiable, repeatable conditions, it produces the opposite.
Cognition, structure, and trust are not three separate problems. They are one problem, each element sustaining the others, and any serious response must address them together. The first step toward seeing through the blind spot is abandoning the belief that if something were wrong, you would already know. The second is building institutions that do not depend on that belief. The blind spot cannot be eliminated. It is too deeply woven into how human beings perceive, trust, and organize themselves. But it can be narrowed, institutionally and personally, by people who understand its architecture well enough to refuse its comforts. That effort, difficult and never complete, is the only reliable defense we have.
Author’s Note
The author served as Chief Corporate Strategy Officer and a director of First Brands Group, one of the cases examined here. He is a defendant in civil proceedings arising from the company's bankruptcy. Criminal proceedings involving the company's principals remain pending and unresolved. The conclusions of this essay do not depend on the outcome of either. Where it addresses cases still in progress, the essay relies only on what has been publicly reported as of the date of publication. The author makes no independent allegation of fraud against any person discussed in this essay; descriptions of pending matters summarize allegations made in public proceedings or reporting and do not state or imply a finding of guilt.