By Terence Tsai

“The CEO is often the most ‘in the dark’ person in the company.” While this statement may come across as ironic, it uncovers a harsh reality in organisational management. Why do CEOs consistently find themselves at the tail end of truth in the organisations they lead? Here, CEIBS Associate Professor of Strategy and International Business Terence Tsai dissects this dilemma from a management perspective, and proposes a solution to break this cycle.

In 2015, the Volkswagen emissions scandal shocked the globe when the carmakers’ engineers were discovered to have installed deceptive software in diesel vehicles to cheat on emissions tests, and to have successfully hidden it from top management for seven years. It wasn’t until the US Environmental Protection Agency (EPA) intervened that the company’s then CEO Martin Winterkorn, appearing bewildered at a press conference, admitted, “I’ve only just found out.”

This isn’t an isolated incident. From Wells Fargo & Co’s fraudulent accounts to fatal flaws in the Boeing 737 MAX, countless cases show that CEOs often end up being the last to uncover the truth about serious issues within the companies they lead. Why are those at the top of the corporate hierarchy often the most likely to be misled? If even the CEO falls victim to deception, can a company really be trusted at all?

From the standpoint of a diligent CEO, lofty ideals such as the building of a top-notch team, the implementation of a thorough reporting process, and regular visits to the grassroots level of the company can count for little in the face of this harsh reality: team cohesion can sometimes dissolve into unspoken agreements, flawless procedures may filter out critical voices, and what is seen during on-site visits may be a carefully prepared "model play".

Why do even effective CEOs fall for this? It isn’t just about simple deceit, nor is it a matter of morality or even competency; rather, it underscores an inherent systemic issue within organisational dynamics.

The Trap of Delegated Agency

The concept of Agency Theory offers a plausible explanation. According to this theory, every organisation can be viewed as a network comprising principals and agents. In this dynamic, principals (such as shareholders) entrust agents (such as CEOs) to act on their behalf. Subsequently, CEOs delegate decision-making authority to employees at various levels through the management hierarchy. While CEOs serve as senior agents at the corporate level, they also rely on other internal employees - their subordinates, who assume new agent roles within this framework. Theoretically, each tier of agents should diligently pursue the principal's best interests, but inherent and challenging conflicts persist.

Firstly, agents naturally hold operational details that principals find hard to access, creating an information advantage that often leads to selective filtering of vital information during communication. When subordinates, driven by self-interest, opt not to fully disclose the reality to the CEO, it can lead to a misinterpretation by the CEO regarding the company’s overall status and the external environment. The fundamental problem here is that subordinates may believe that concealing information can avert short-term conflicts or unfavourable consequences, yet in the long run, such conduct may undermine the company’s overall interests.

Secondly, behavioural deviations arising from divergent incentive objectives. Shareholders pursue long-term value, management prioritises short-term results, the board of directors focuses on strategic growth, and the executive level is concerned with personal performance metrics. This fundamental misalignment of objectives leads to decisions by agents at different levels conflicting with the organisation’s holistic interests. When achieving quarterly targets yields more immediate rewards than upholding compliance standards, the rational choice for employees may lean towards information concealment or even deceitful actions.

Finally, when information asymmetry and conflicting incentives reinforce each other, a multiplier effect of moral hazard is created. Agents not only hold an advantage in information but also face a distorted incentive structure, leading to the systematic rationalisation of actions that harm the principal’s interests. This dual dilemma makes information distortion within organisations not a random occurrence, but an inevitable outcome of institutional design flaws.

The Volkswagen emissions scandal serves as a vivid illustration of this predicament. The engineers were not inherently deceitful but, operating within an evaluation system that rewarded meeting standards and penalised violations, they pragmatically opted for the path of using cheating software - because for the agents, pressure to keep their jobs was far more tangible than pressure to uphold environmental commitments. Junior employees identified issues but feared impacting team performance, while middle managers were informed but worried about their career progression, ultimately leading to senior leadership being completely blindsided. This prolonged collective silence which lasted for seven years, was only shattered when the EPA conducted third-party testing, starkly affirming the core proposition of agency theory: in the presence of information asymmetry and conflicting incentives, concealment will become the norm in organisations.

Distorted Information Transmission

In an organisation, the accuracy of information transmission is a crucial factor affecting the quality of decision-making. While the CEO is supposed to be the central figure receiving comprehensive information as the top manager of the company, the reality often paints a different picture - the CEO is typically the last to know about the true situation within the organisation.

In an ideal scenario, information would freely circulate within an organisation. However, in reality, communication within organisations is rarely straightforward or transparent. It often undergoes a hierarchical dissemination process, passing through filters and subjective interpretations. As highlighted by management experts James March and Herbert Simon in their book Organisations, members of an organisation tend to simplify and screen information based on their limited rationality during the communication process. This often results in senior decision-makers receiving incomplete or even deceptive information.

The hierarchical structure of organisations further exacerbates the problem of information distortion. In a multi-layered organisational setup, information must navigate through several intermediary levels before reaching the CEO. With each layer it crosses, there is a chance that the information will be further screened or modified. The more layers there are, the greater the risk of information distortion.

In most traditional enterprises, the organisational structure takes on a classic pyramid shape. The CEO sits at the very top, while information must climb up from the lowest levels, passing through each tier. Each level has the potential to function as an “information filter” or “distortion mechanism.” Management expert Henry Mintzberg, in his work The Structuring of Organisations, offers a systematic classification of organisational structures, and points out that in a typical “machine bureaucracy” structure, information is required to move through the hierarchy. While this structure promotes procedural consistency, it also tends to cause communication delays and information distortion.

Moreover, as information traverses through layers of communication, it undergoes multiple rounds of “language translation.” Middle managers often interpret the actual situation into formats that are easily digestible for senior executives - such as KPIs, reports, charts, and indicators. Throughout this process, the depth and intricacy of the information are significantly watered down. By the time the information reaches the CEO, it no longer represents the authentic reflection of the original events or problems, but rather “abstract data points” that have been reinterpreted multiple times. The CEO is presented with outcomes, not the process; figures, not the narrative.

Adding to the complexity is the fact that information is not just simplified in form but also “positioned” in content. Driven by their individual interests, different departments may offer starkly different interpretations of the same issue. The sales department may play down customer complaints and highlight market potential; the technical department may prioritise system security while minimising progress concerns; the finance department may emphasise cost control while overlooking operational challenges. As a result, the information presented to the CEO is often a “compromised version” following negotiations between various departments, rather than an accurate representation of the actual problem at hand.

For example, the Boeing 737 MAX’s flight control system MCAS (Maneuvering Characteristics Augmentation System) had critical safety issues that ultimately resulted in multiple air crashes. However, before these accidents, the top management of the company did not fully grasp these risk factors. Why did this happen? The organisational hierarchy at Boeing was complex, leading to a significant communication gap between frontline engineers and top management. The middle management either failed to, or possibly chose not to, escalate the problems. It wasn’t until after the accidents occurred that the CEO truly realised the gravity of the situation. This was not just a technical problem but a classic example of failed organisational information dissemination.

Within organisations, information is not a free-flowing stream. When information is filtered, distorted, or even concealed layer by layer, even the most powerful CEO can only grasp the truth afterwards. If an organisation fails to address the structural issues in information flow, a CEO can never truly “have a handle on the whole picture.”

Psychological Safety Deficit

Organisational culture serves as the invisible hand within a company, shaping the daily behaviours, communication styles, and value assessments of employees. In numerous enterprises, a culture that prioritises compliance, pursues performance, and avoids conflicts reigns supreme. Within this culture, employees are not always inclined to voice the issues they are aware of. Even when they recognise risks in a process or errors in a decision-making direction, they frequently opt for silence.

This phenomenon is not a reflection of individual timidity, but rather the unconscious influence of organisational culture shaping a norm of sharing only good news and no bad news. Even with CEO encouragement of open communication, employees may still question the safety of speaking up. Superficial open communication masks a deeper organisational silence - a systematic reluctance to speak, inquire, or convey crucial information.

Why do employees opt for silence? The key lies in the impact of organisational culture on individual behavioural choices, which leads to the concept of “psychological safety.” Introduced by Harvard Business School professor Amy Edmondson, the concept of psychological safety highlights individuals feeling secure within a team to freely express authentic viewpoints, admit mistakes, and pose questions without the fear of humiliation, rejection, or punishment.

In the absence of psychological safety, employees, despite being aware of existing issues, are hesitant to take the risk of revealing the truth. They worry that speaking out may lead to being perceived as disloyal or unprofessional. This fear is especially pronounced in strict hierarchical systems or high-pressure performance cultures. As a result, information is self-censored before it is shared, and the truth silently fades away in the minds of frontline staff, leaving the CEO in the dark.

It can be argued that organisational culture represents external ambiance, while psychological safety embodies the internal experience. When an organisation lacks a culture of inclusivity, forgiveness, and attentive listening, the establishment and preservation of employees’ psychological safety becomes challenging. Even if systems like suggestion boxes, anonymous feedback channels, or transparent reporting mechanisms are in place, if employees do not genuinely believe that speaking the truth is safe, these frameworks end up being mere symbolic gestures and fail to truly encourage open communication. Therefore, employee silence is often not due to ignorance or indifference but rather due to fear and distrust. This is not a matter of information deficiency but rather a breakdown in trust.

The truth is not unknowable but rather unspoken. Organisational silence is more perilous than misinformation because it breeds a deceptive sense of security, making top management mistakenly believe that “no bad news is good news,” ultimately leading them into strategic blind spots.

The Wells Fargo & Company’s fake accounts scandal is a vivid example of the detrimental consequences of collective organisational silence and the lack of psychological safety. To meet unrealistic sales targets and achieve so-called “cross-selling” goals, numerous employees opened fake accounts for customers without authorisation over several years. Despite some employees recognising the risks associated with this behaviour and attempting to alert management, their voices were not taken seriously and were quickly suppressed within an organisational culture that was highly performance-oriented, had a strict punishment system and lacked psychological safety. Many employees admitted later that they refrained from questioning their set targets out of fear of being perceived as “not ambitious enough” or “underperforming,” thereby facing the risk of demotion or dismissal. As a result, the issue remained silent within the organisation for years until it was exposed by the media and investigated by regulators.

The phenomenon of information filtering may appear, therefore, to be a structural issue on the surface; but, at its core, it’s a cultural one. While it may seem like information is being passed up the hierarchy, in essence, it involves a process of “interpreting upwards” and reviewing at each level. At each level, information is being processed- not to convey facts more accurately, but to align with the preferences of superiors more safely. This is not a result of individual moral decline, but rather a form of “survival wisdom” that has subtly developed within the organisational culture. Ultimately, what reaches the CEO is no longer authentic information but a customised message that caters to superiors’ expectations and deviates from the facts in order to conform to organisational tendencies.

The culture of “superficial democracy” and obeying superiors is the underlying cause of information distortion within organisations. Without confronting this “cultural distortion,” even with robust systems, efficient processes, and abundant data, the CEO will still be unable to truly see the truth.

The Way Forward

To address this persistent organisational issue, it is imperative to shift away from individual oversight and reconsider at a system level why information fails to effectively reach the top levels of an organisation. The root of this problem does not lie in whether executives conduct site-visits, nor whether the team is loyal and diligent, but rather in whether the organisation fosters a culture of truth circulation.

First and foremost, as revealed by agency theory, every employee within the organisation serves as a form of “agent.” The inclination for agents to resort to concealment often arises from the distortion of incentive mechanisms. When short-term KPIs override everything else, and when sharing good news is more rewarding than reporting bad news, silence and covering up become the rational choice for employees.

For the organisation to alter this paradigm, they must fundamentally adjust their incentive systems, incorporating long-term thinking into performance frameworks and incentivising problem identification and exposure of risks. Agents are only inclined to act as whistleblowers rather than cover-up when personal gains strongly align with the interests of the organisation.

Secondly, organisations need to change the way information flows. If the CEO can only rely on traditional, singular sources of information, then what they see is merely a modified version of reality. Therefore, organisations must break this one-way, top-down logic of information reporting and establish diverse information loops. Only by creating multiple pathways for truth to emerge can the CEO hear the unfiltered voices from the front lines.

Nevertheless, what truly determines whether the truth can be heard in an organisation still lies in the foundational values of its organisational culture. While systems can be designed and processes optimised, the courage of employees to speak the truth depends on whether they can feel psychological safety. When an organisation tacitly or even actively promotes flattery and catering to superiors' preferences, employees naturally perceive speaking the truth to be a high-risk behaviour.

Only when employees believe that telling the truth will not lead to humiliation, rejection, or punishment are they likely to speak up. Psychological safety is not built on slogans, but is gradually nurtured through daily organisational practices. Leaders must set an example by admitting their ignorance and mistakes, openly embracing unwanted points, and consistently signalling that problems are opportunities for improvement, not threats. It is only when every individual in the organisation understands that voicing dissenting opinions carries no negative consequences that the truth will be voiced, and information will flow freely. This is how an organisation can truly become a place where honesty is valued, listened to, and embraced for the sake of growth and progress.

In today’s organisations, AI technology is silently transforming the way information circulates, offering new opportunities to decode the CEO’s blind spots.

In traditional organisations, the acquisition and dissemination of information heavily relies on human judgment and interpersonal networks, inevitably influenced by subjective consciousness, hierarchical frameworks, and power plays. The emergence of AI is shattering this paradigm. Through technologies like natural language processing, behavioural analysis, and emotional recognition, AI can autonomously identify abnormal signals and potential risks from intricate data streams without relying on hierarchical intermediaries.

Take, for instance, frequent searches for a specific clause in internal systems by employees, repeated complaints by customers on social media, or atypical communication patterns on team collaboration platforms - these once-overlooked pieces of information can surface through AI-powered analysis, providing management with an unparalleled level of “organisational perceptual capability.”

Certainly, AI is not a cure-all, as data does not inherently transform into insights, and it cannot replace the judgment and accountability of leaders. Nevertheless, it is undeniable that the rise of AI is driving organisations to rethink how information should be managed and presenting CEOs with a fresh realm of possibilities.

The true challenge does not lie in the further sophistication of technology, but rather in the organisation’s willingness to utilise it to uncover the truth. AI has the potential to serve as the CEO’s “second pair of eyes,” but only if the CEO desires to confront the less flattering aspects of their organisation. The key also lies in whether they are willing to establish an organisational ecosystem that encourages human expression while embracing technological insights.

A CEO cannot know it all, but they can be the one who embraces the truth. They do not need to grasp every detail, but they must ensure that, at critical moments, the voices least inclined to be heard are given a chance to speak.

The essence of leadership lies not in having complete control over information, but in fostering an environment in which truth can emerge. A genuinely mature organisation is not one without problems, but rather a place where problems can be promptly identified, openly addressed, and collaboratively fixed.

Terence Tsai is an Associate Professor of Strategy and International Business at CEIBS. His research interests include multinational corporations, sustainability strategy, environmental management, organisational theory (environmental adaptive theories) and Chinese management.