Is algorithmic management creating a new hierarchy in the workplace?, ETHRWorld

Time of India

23-05-2025

Highlights AI surveillance is reshaping workplace power dynamics.
Employees may engage in 'productivity theatre' to impress algorithms.
High monitoring can stifle creativity and reduce meaningful output.
Advt
Advt
What amount of wokeism is permissible in the workplace?
Increased social awareness and activism, i.e., wokeism, among employees and its interaction with a company's culture has left organizations puzzled about how much of it to encourage so as to preserve operational stability. See More Details
By & ,
ETHRWorld Contributor
Join the community of 2M+ industry professionals Subscribe to our newsletter to get latest insights & analysis.
Download ETHRWorld App Get Realtime updates
Save your favourite articles
Scan to download App
A recent Bartleby column in the Economist satirizes AI surveillance in team meetings, analysing body language, eye-rolling, private messages, and even lip-reading when cameras are off, touching on the unsettling reach of workplace monitoring technology.A Wall Street Journal podcast reveals companies deploying software to identify promotion candidates and flight risks, analyse chat logs between conflicting employees, and increasingly adopt HR chatbots like paidleave.ai.Many executives are pushing departments to 'do more with less' via the adoption of AI to evaluate performance, predict attrition, allocate bonuses, resolve conflict, and even drive promotions and terminations. Algorithmic oversight now extends far beyond traditional task metrics into deeply personal realms.These developments may open a Pandora's box. As workplaces become algorithmically optimized, a subtle but profound shift in organizational power dynamics is bound to shape up. This transformation warrants deeper scrutiny not merely as a technological trend but as a fundamental restructuring of how power is experienced, expressed, and executed in the workplace. After all, power structures form the bedrock of organizational behaviour.While software to automate managerial functions is not new, its modern successors are special.Algorithmic Affect Management ( AAM ) combines good old workplace management with affective computing to create systems that monitor, analyse, and respond to workers' emotional states and reactions. AAM captures data on everything from tone of voice and facial expressions to brain activity and physiological responses to inform automated management decisions about productivity, performance evaluation, and resource allocation.Companies are extensively taking to algorithmic management . OECD's Employer Survey 2025 says 90% of US firms use at least one AAM tool to instruct, monitor, or evaluate workers. European countries have an average adoption of 79%. According to a McKinsey study, the use of AI by organizations in one or more business functions has increased from 55% in 2023 to 78% in 2025.On one hand, the ecosystem is ripening to adopt more AAMs, while on the other, their hair-splitting abilities are coming into full view.Organizational hierarchies have long depended on an unspoken grammar of submission—deferential gestures, seating arrangements, and even the cadence of laughter in meetings. These cues, as organizational psychologist Robert Sutton observed in The No Asshole Rule, are often subtle reinforcements of status and authority, echoing animalistic behaviours of dominance and deference.But these old cues are being replaced. The rise of 'dystopian' employee productivity monitoring software has created a new power dynamic where algorithms, not humans, determine status.When real-time productivity dashboards and behaviour analytics determine employee value, work becomes a performance calibrated to machine perception.Prof Lindsey D Cameron of the Wharton School has found elements of 'deviance tactics' by ride-hailing workers to align their efforts with managerial interests. We may imagine analogous white-collar employee behaviours to game human-algorithm interactions when faced with red notices, or even otherwise, for falling below arbitrary productivity or 360-degree alignment thresholds.Employees are increasingly poised to navigate and, at times, manipulate AAMs. These systems often rely on data from emails, keystrokes, typing speed, backspace frequency, email volume, and other digital interactions to create 'productivity graphs' that compare workers against their peers on sometimes vague success-driven parameters.According to research, more than one in ten workers will actively attempt to deceive AI-driven tracking systems. To maintain a favourable profile, employees may engage in 'productivity theatre,' such as scheduling emails to be sent during non-working hours or frequently updating collaborative documents, creating an illusion of constant activity. Others might use tools to simulate keyboard activity or remain logged into communication platforms to appear available to align with the algorithm's expectations.Some workers may use more sophisticated tricks like VPNs or multiple devices to avoid surveillance and separate monitored from unmonitored tasks. Reports suggest that companies that sell these monitoring tools and provide AI automation services are often the same, suggesting a troubling pathway from surveillance to replacement, almost encouraging one to trick them.By learning how algorithms evaluate them, workers adapt their behaviour to meet metrics without real engagement. This dynamic creates a growing awareness among white-collar workers of the surveillance capabilities of AAMs and a proactive approach to exert autonomy and retain control over their work environment.Such behaviour stands to aggravate as companies begin to use more powerful AAMs. AI is introducing a new layer of hierarchy in organizations and associated inconveniences. A new area of people management is emerging with it.The Institute for the Future of Work (IFOW) says worker reactions to AAM are largely negative. Its Data on our Minds: Affective Computing at Work research found that fewer than 10% agree that these systems positively impact their wellbeing, safety, or health. The psychological toll is significant. Most employees report that AAM forces them to work faster, meet tighter deadlines, and change their working habits, with 29-34% experiencing increased 'techno' stress levels.If AI is already generating up to 30% of code at Microsoft, we may extrapolate and attribute one-third of an employee's yearly appraisal directly to AI, too. There is also the perception that the bosses will tighten based on these metrics: based on what the algorithm's 'expectations' from an employee are.These are not simply tools that assist human managers—they're increasingly becoming managers themselves, fundamentally altering workplace dynamics and creating a paradoxical leadership vacuum.While algorithms efficiently distribute tasks, monitor performance, and enforce standards with relentless consistency, they simultaneously diminish the human elements of management that employees routinely rank as most valuable: empathy, contextual understanding, and, in particular, interpersonal trust.Moreover, algorithms risk being able to evaluate without being evaluated. About two-thirds of managers in recent OECD surveys are worried about the unclear accountability of AAMs for wrong decisions and being unable to follow the tools' decision-making logic. While this will resolve enroute to Artificial General Intelligence, the intervening time could be one of chaos.The invisible hand of AAMs is simultaneously more present through continuous monitoring yet more absent in meaningful human interaction. The International Labour Organization has voiced managers' concerns about diminishing interpersonal skills as their roles become increasingly focused on interpreting and implementing algorithmic recommendations rather than exercising independent judgment.Such technological mediation creates a black box so that manager-employee relationships become increasingly opaque and transactional, feeding employee disengagement despite unprecedented levels of connectivity.Prof Ifeoma Ajunwa of Emory Law School notes in her research on Automated Governance that such black boxes could potentially reinforce entrenched power structures. For example, the 'old boys' club,' where male employees gain advantages through privileged networks—a dynamic particularly pronounced in India's hierarchical corporate environments, where caste, regional, and linguistic biases often operate beneath the surface.Perhaps the most surprising development in the algorithmic workplace is what Prof Erik Brynjolfsson of Stanford University has identified as the 'modern productivity paradox '. Despite unprecedented capabilities for monitoring and optimizing every aspect of work, from keystrokes to brain waves, many organizations implementing comprehensive AAMs are experiencing stagnant or declining productivity when measured by meaningful output rather than activity metrics. This counterintuitive finding challenges the fundamental premise of algorithmic management.One Reddit user addresses the heart of the problem: 'The better I got at looking busy, the less time I had to think.'When workers know they are constantly monitored down to their neurological activity, many respond by directing their ingenuity toward satisfying algorithmic metrics rather than solving substantive problems. It echoes economist Charles Goodhart's claim that a measure ceases to be effective when it becomes a target.Neuroscience studies reveal that the brain's default mode network, essential for creative thinking and novel connections, becomes suppressed under conditions of surveillance. Resources shift to the brain's executive control network, optimizing for predictable performance rather than creative thinking.A workplace surveilled by AI may appear highly active yet may underperform as regards adaptive innovation and problem-solving. These are precisely the human capabilities organizations need most in a rapidly changing AI-powered economy to realize, as Sundar Pichai recently said, an 'explosion in knowledge, learning, creativity, and productivity'.The path forward requires a thoughtful integration that preserves human agency and creativity while leveraging AI's analytical strengths. Research by the Foundation for European Progressive Studies suggests that organizations achieving this balance share a few key characteristics.First, algorithmic transparency is critical. Employees should understand how algorithms work, not just that they are being used. This boosts trust, reduces anxiety, and limits gaming behaviour. Second, clear human override protocols are essential. Final decisions must remain with human managers who can factor in nuance and context. Third, organizations should invest in AI sandboxing to test AAMs before plugging them into the company's culture. Sandboxing is useful from a regulatory standpoint as well.Organizations must actively foster self-efficacy in AI learning. Research on South Korean professionals found that employees who feel confident in their ability to understand and work with AI experience significantly less stress during adoption.Crucially, the goal should be augmentation over automation: AI should enhance human capabilities at all organizational hierarchies, not replace judgment. This allows managers to focus on coaching and interpersonal leadership.Organizations must also embed metrics of trust, such as transparency, explainability, fairness, reliability, and privacy, into AI design and implementation. These must be communicated effectively to users, including non-experts, to ensure perceived trustworthiness matches actual performance.Beyond the scope of a company, regulations and standards overseen by governments are needed to calibrate trust and ensure ethical usage across diverse cultures. Dynamic, context-sensitive trust models, especially in AI–AI interaction, remain a critical research frontier.Finally, organizations must aspire to a work culture that values human ingenuity, encouraging reflective thinking and autonomy. Balancing efficiency with cognitive freedom is vital. Only by harmonizing human and machine intelligence can organizations unlock AI's transformative potential without compromising the human core of work.