The defeat device worked. It correctly identified when a vehicle was being emissions-tested and activated clean-mode. In normal driving it deactivated, emitting up to 40 times the legal limit of nitrogen oxide. This was not a rogue act by a junior engineer. It was a governance system that made it impossible for the truth to travel upward — and a leadership alignment that made it rational to keep it silent.
What Actually Happened
The mechanics of Dieselgate are now well-documented. Starting in 2006, a small group at Volkswagen's Wolfsburg headquarters — facing a CEO target of becoming the world's largest automaker by selling over 10 million cars annually — confronted a problem that their engineers could not solve cleanly: diesel engines could meet US emissions standards in a test environment, or they could perform well in real-world driving. They could not do both.
The defeat device was the solution. Software in the engine control unit detected when the vehicle was being tested — measuring steering angle, speed, and whether the windshield wipers were running — and activated full emissions controls. On the road, those controls switched off. The result was a car that passed every regulatory test and produced up to 40 times the permitted level of NOx in normal use.
2006: Defeat device first deployed across VW's TDI diesel range as a response to an unresolvable engineering conflict between performance and US emissions compliance.
2007: According to the SEC complaint, CEO Martin Winterkorn and other senior VW executives were made aware of the defeat device at a meeting with engineers discussing emissions problems with "clean diesel" vehicles.
2009–2015: The defeat device runs through six consecutive US model years, covering 500,000 vehicles in the US and approximately 11 million globally. Der Spiegel later reported that at least 30 people at management level knew of the deceit.
2013: The International Council on Clean Transportation commissions real-world emissions tests of diesel vehicles sold in the US, finding on-road NOx levels dramatically exceeding lab results.
2015: The EPA issues a Notice of Violation. VW initially denies and then admits the defeat device. CEO Winterkorn resigns. The cover-up had held for nine years and six model years.
2017: Volkswagen pleads guilty to criminal charges — conspiring to defraud the US government, obstruct justice, and engage in wire fraud. The agreed Statement of Facts records that management asked engineers to develop the defeat devices because its diesel models could not pass US tests without them — and deliberately sought to conceal their use.
The financial consequences built over years. VW announced plans in 2016 to spend €16.2 billion on the scandal. The criminal fine alone was $2.8 billion — the largest ever for an automaker in the US. The total cost of fines, penalties, financial settlements, buybacks, and vehicle refits reached $38 billion and counting — approximately half the company's market capitalisation at the time of discovery.
"Our company was dishonest with the EPA, and the California Air Resources Board and with all of you." The CEO of Volkswagen of America said this publicly in 2015. It had been true since 2006. Nine years of dishonesty that every governance structure Volkswagen had was designed — unconsciously — to protect.
The Governance Anatomy — Why No One Said Anything
The question that matters for governance is not why engineers built the defeat device. Under the pressure they faced, the reasoning was traceable. The question is why no one in nine years — not a lawyer, not an auditor, not a risk officer, not a board member — ever successfully escalated it through a governance channel that changed the outcome.
The answer lies in how VW's organisation was structured and what that structure made rational. VW under Ferdinand Piëch and then Martin Winterkorn was explicitly, publicly autocratic. Piëch boasted that he managed by "terrifying his engineers." Former employees described a workplace in which subordinates were afraid to admit failure or contradict superiors. Company leaders bullied employees. The goal — ten million cars sold, global leadership wrested from Toyota — was not a strategic aspiration. It was a personal mandate from the top, communicated downward with force.
In that environment, middle managers who knew about the defeat device had an entirely rational calculation to make. Escalating meant delivering bad news in an organisation where bad news had career consequences. Staying silent meant being part of a scheme that had been running for years across management levels — and that therefore carried implicit protection from the top. The information did not flow upward not because systems failed but because the systems were operating exactly as the culture had designed them to.
The Jones Day internal investigation — conducted after the scandal broke and provided to the Department of Justice to secure a reduction in VW's criminal fine — was never made public. German authorities had to search the law firm's Munich offices to seize the report. Whatever it found about who knew and when, the structure of the concealment was clear: an organisation in which bad news was structurally unreachable by the board, and in which the compliance function had never developed the independence or the mechanisms to surface what the engineering function already knew.
The NAVETRA™ Domains That Were Failing at Volkswagen
NAVETRA™ measures the ten organisational and human domains that determine whether governance actually functions. It does not measure whether a company has compliance frameworks, audit committees, or ethics policies — VW had all of these. It measures whether the people and structures responsible for those frameworks are aligned, capable, and positioned to make them work in practice. At VW, five domains were failing simultaneously — and they were failing in ways that were not only predictable but structurally overdetermined by the culture Piëch and Winterkorn had built.
Are the CEO, supervisory board, and management board operating from the same shared picture of reality — or is critical information being stopped before it reaches the people with authority to act?
At VW: the management board was briefed on the defeat device years before the EPA notice. The supervisory board — which included Ferdinand Piëch, the architect of the culture that made the fraud rational — never received that information through a governance channel. Leadership Alignment at VW was not absent: every function was intensely aligned to Winterkorn's ten-million-car target. What was absent was alignment around the truth. The organisation was aligned to a story, not a reality — and the gap between them was $38 billion.
Does the organisation have the internal ownership structures, escalation paths, and process discipline to surface compliance risk before it becomes a regulatory or criminal event?
At VW: the defeat device was known across engineering, environmental compliance, and US regulatory affairs teams. A compliance function with genuine escalation paths — including paths that reached the supervisory board independently of executive management — would have surfaced it. VW's compliance function was structurally subordinate to the executive team it was supposed to monitor. When Oliver Schmidt, the executive in charge of VW's US environmental compliance, was informed of the defeat device, he did not escalate to a compliance channel. He was sent to convince US regulators to allow the 2016 model year to proceed.
Are engineering, legal, compliance, and executive leadership operating from shared risk visibility — or are critical findings contained within functions, rationalised locally, and never reaching the people with authority to act?
At VW: the defeat device was known across multiple functions simultaneously for years. At no point did engineering, legal, regulatory affairs, and executive leadership hold a shared discussion of what the organisation was doing and what it meant. Each function processed its knowledge locally, within the incentive structure it faced. Cross-Functional Alignment does not require conspiracy to fail — it requires only that each function has a rational reason to keep what it knows within its own boundary. At VW, every function did.
Is the organisation structurally aligned to its stated values — or are the actual incentive structures, cultural norms, and decision-making patterns pointing in a different direction?
At VW: the organisation was aligned to output. Volume. Deadlines. The ten-million-car target. Every promotion, every bonus, every signal about what mattered in VW's culture was aligned to that goal — and to the implicit understanding that obstacles to it were not to be escalated upward. The organisation's stated values — integrity, transparency, compliance — were performative. The real alignment was to delivery at any cost. Organisational Alignment failure does not look like a sudden collapse. It looks like nine years of engineers who knew something was wrong continuing to show up to work.
Is institutional knowledge about compliance obligations, legal risk, and governance requirements travelling reliably across functions, geographies, and leadership transitions — or is it siloed within functions that have no shared accountability?
At VW: the regulatory risk associated with the defeat device — what it meant under US Clean Air Act law, what the penalty exposure was, what the disclosure obligations were — was almost certainly understood by legal and regulatory affairs. That knowledge never reached the supervisory board in a form that triggered action. This is not a gap in information. It is a gap in the organisational architecture that determines who shares what with whom, under what conditions, and with what consequence if they choose not to. At VW, that architecture was built to protect the target, not the truth.
The Scale of What Organisational Failure Cost
Separating what VW genuinely could not have controlled from what its own governance caused is difficult — the defeat device was entirely a governance failure, with no exogenous component. But the financial decomposition is instructive:
VW's market capitalisation lost €27.4 billion in value within five days of the EPA notice. The criminal fine of $2.8 billion was the largest ever imposed on an automaker in the US. Seven current and former employees faced criminal charges. CEO Winterkorn — who the SEC asserts was personally briefed on the defeat device as early as 2007 — was indicted in the US in 2018 and faced criminal trial in Germany in 2024.
The compliance architecture VW built after Dieselgate — including a court-appointed independent compliance monitor who supervised the company for three years — is a $38 billion proof of concept for what a functioning NAVETRA™ deployment should have looked like before the EPA notice arrived.
$38 billion. Nine model years. 11 million vehicles. Scores of employees who knew. Not one governance channel that worked. Dieselgate is not a story about what engineers did. It is a story about what organisations do when Leadership Alignment points toward a target, Organisational Alignment enforces silence, and Internal Risk Management has no path to the supervisory board that doesn't run through the executive team it is supposed to monitor.
The signals were not absent. They were structurally unreachable — until the regulator knocked on the door.
The Question for Your Organisation
Every organisation that has ever set an ambitious target and built a culture around achieving it carries some version of this risk. Not the specific risk of a defeat device — but the structural risk of an organisation in which bad news travels slowly, in which middle managers have rational reasons to filter what goes upward, and in which compliance functions are positioned to report to the executives they are supposed to monitor.
NAVETRA™ does not measure whether your compliance policy is well-written. It measures whether your organisation is actually aligned to make it work — whether the people responsible for surfacing risk have the structural independence, the cultural safety, and the escalation paths to do so before the regulator, the whistleblower, or the investigative journalist gets there first.
At Volkswagen, the governance architecture was not absent. It was performing exactly the function the culture had optimised it for. That is the most important thing Dieselgate teaches — and the most dangerous thing to miss.
Execution risk that isn't measured doesn't disappear. It accumulates — until the accounting forces you to see it.
