^{“Would I rather be feared or loved? Easy – both. I want people to be afraid of how much they love me.” Michael Scott, The Office}

Some of the clearest illustrations of dysfunctional leadership don’t come from case studies or court filings, but from fiction. In The Office, Steve Carell’s Michael Scott is remembered as a well-meaning but deeply insecure manager – desperate to be liked, terrified of being overlooked, and chronically anxious about his own relevance. What makes the character enduring is his instinct to ensure that he always remained the centre of attention.

Michael Scott routinely undercut his team to reinforce his own importance. He interrupted presentations, hijacked credit for others’ ideas, and turned collective achievements into personal performances. Talented employees were tolerated only so long as they did not threaten his status. When they did, he deflected, belittled, or reasserted control – often unconsciously.

Played for comedy, the behaviour is uncomfortable because it is familiar. Many organisations have encountered their own version of Michael Scott: managers whose authority depends less on enabling others and more on making sure no one outshines them. The result is rarely dramatic confrontation. Instead, it is quieter – a culture where people learn to hold back, where speaking up feels risky, and where initiative is carefully rationed.

Habits of Leadership at Harvard Business School

New research from Harvard Business School suggests this dynamic is not just anecdotal or fictional. It is a recurring pattern in real organisations, driven by status anxiety and reinforced by hierarchical systems. And as artificial intelligence begins to challenge long-held assumptions about expertise and authority, the same instinct that drove Michael Scott to dominate the room is reappearing – this time directed not only at people, but at the technologies that threaten to outshine their managers.

In the ideal workplace, managers are champions of their team’s growth: they nurture talent, delegate with trust, and create conditions for innovation. Yet a growing body of research reveals a paradox at the heart of many organizations: some managers undermine their most capable employees for fear of being outshined and losing their grip on power. This dynamic, documented in recent Harvard Business School research, bears striking similarities to how some leaders resist new technologies like artificial intelligence (AI), even when those tools could benefit the organisation. 

Top-Down Sabotage: When Managers Undermine Talent

A working paper from Harvard Business School researchers Hashim Zaman and Karim Lakhani shines a spotlight on a toxic yet underexplored phenomenon: top-down sabotage, where managers actively impede the career progression of their high-performing subordinates. In a survey of 335 executives across industries, they found:

• About 30% of respondents reported observing sabotage in their organization, and more than 70% had witnessed such behavior at some point in their careers.
• Around 28% said they had been victims of managerial sabotage.
• While financial motives played a minor role, status concerns – the fear of losing power or being outcompeted – were a primary driver of these behaviors.

Zaman explains that in highly competitive workplaces, ambition and self-preservation can trump collaborative goals. Managers may see a talented employee not as a resource to develop but as a future peer, and potential threat. In response, the manager might withhold opportunities, obscure credit for work, or consciously shape evaluations in ways that slow a rising employee’s trajectory. 

This insight resonates with classic research on “toxic leaders” and workplace undermining, where insecurity and a craving for power can cause leaders to damage their organizations as they seek to consolidate their own position. 

The Organisational Impact: A Culture of Fear and Distrust

When managers block the success of talented staff, the consequences extend far beyond individual careers:

• Talent attrition increases as ambitious employees choose to leave environments where their contributions aren’t valued.
• Psychological safety declines, making teams less willing to share ideas or take risks – a critical ingredient for innovation.
• Organizational performance suffers when promising people are stifled rather than supported.

Business culture expert Amy Edmondson’s extensive work on psychological safety underscores that high-performing teams require environments where people can speak up, experiment, and be vulnerable without fear of negative repercussions. In high-trust cultures, employees are more likely to collaborate and challenge the status quo – exactly the opposite of what thrives under sabotage.  

Combating top-down sabotage requires transparent evaluation systems and incentives aligned with collective success, not zero-sum performance comparisons. According to the Harvard research, organizations that minimize subjective managerial discretion in reviews and cultivate open communication were less prone to sabotage dynamics.  

The AI Angle: When Managers Resist Innovation

“People will never be replaced by machines,” says Michael Scott. “In the end, life and business are about human connections. And computers are about trying to murder you in a lake. And to me the choice is easy.”

Interestingly, similar patterns appear when organisations grapple with technological change. Just as some managers feel threatened by high-performing subordinates, other leaders resist adopting artificial intelligence, not always due to technological limitations, but because of psychological and cultural factors.

Resistance to AI in the workplace often stems from perceived threats to job security, loss of control, and fear of the unknown, according to recent research. Employees and leaders alike may react defensively when new tools disrupt familiar workflows or diminish traditional sources of expertise and authority.

Research on AI resistance by Ismail Gölgeci at Aarhus University and Brad McKenna at Norwich Business School notes that resistance to technology, including AI tends to peak when individuals feel their roles are vulnerable, they lack understanding, or the organisational culture doesn’t support experimentation and learning. 

In this sense, the emotional drivers behind resistance to AI and sabotage of talented staff share a common root: fear of loss – whether it’s loss of status, control, or personal relevance. A manager who sees AI as a threat may respond with subtle undermining behaviors:

• Delaying AI rollouts under the guise of caution.
• Prioritising manual work to maintain traditional hierarchies.
• Downplaying the value of AI insights to preserve their own expertise.

These behaviors mirror how insecure managers might withhold opportunities from talented staff. Not because the tools or people lack value, but because they challenge the existing power structures.

Leadership Lessons: Creating Cultures of Growth and Innovation

So what separates organizations that thrive from those caught in cycles of internal resistance and sabotage?

1. Emphasize Psychological Safety

Leaders should cultivate environments where dissent is welcome and questions are valued – whether about performance processes or the adoption of new technologies. Psychological safety doesn’t eliminate fear, but it channels it constructively.

Research shows teams with high psychological safety are more innovative and resilient, particularly in times of change.

2. Align Incentives with Collective Value

When performance systems reward individual competition over shared success, sabotage and resistance flourish. Structural designs that reward team achievements and transparent metrics reduce the incentive to undermine others or hoard knowledge.

3. Lead AI Adoption with Empathy

Resistance to AI often masks deeper concerns about job security and autonomy. Leaders who communicate openly, demystify AI capabilities, and involve teams in shaping adoption strategies build trust and reduce fear-based responses. 

In fact, emerging leadership practice emphasizes human-AI collaboration rather than human vs. AI. When leaders frame AI as augmenting human strengths, not replacing them adoption accelerates and fear declines.

4. Build Competence and Confidence

Part of resisting technology comes from discomfort. Investing in training and peer learning networks helps demystify AI and shift mindsets from suspicion to curiosity.

At the same time, leadership development should include emotional intelligence training, so managers can distinguish between legitimate concerns and fear-driven resistance.

The True Test of Leadership

“Sometimes you have to take a break from being the kind of boss that’s always trying to teach people things,” Michael Scott philosophically explains. “Sometimes you just have to be the boss of dancing.”

Both managerial sabotage of top talent and resistance to AI reveal a deeper truth about organisational behaviour: the barriers to progress are often human, not technical.

Whether the challenge is supporting your brightest employees or guiding your organisation through AI transformation, the most effective leaders are those who confront fear with transparency, align incentives with shared success, and foster cultures that value trust over insecurity.

Interested in this topic? You might also like this…