Algorithm In The Office: Balancing Automation And Employee Growth

Saurabh Mehta

Strategic Human Resource Management in the Era of Algorithmic Technologies: Key Insights and Future Research Agenda, 2024; Kim, S., Khoreva, V. and Vaiman, V.

Source

Algorithm In The Office: Balancing Automation And Employee Growth

Problem of Practice: Technology is becoming increasingly important in the workplace, as it can help reduce manual routines for human resources (HR) administration, such as payroll. This enables HR professionals to focus on core activities such as talent development. But this gain is accompanied by a paradox. Most HR leaders perceive that spending on technology often fails to add value in key functions such as talent retention, better hiring and stronger productivity. HR leaders are experiencing a vital gap: how to frame guidelines that balance work efficiency with human-centric responsibilities? Recent research by Sunghoon Kim, Violetta Khoreva and Vlad Vaiman outlines a three- lens framework that can improve job design, work structure and HR service delivery, ensuring effective management of technology and workers¹

Tool view of technology

A recent Gartner survey revealed that only 35% of HR leaders believe that their HR technology can help the company achieve its goals.² Only 24% of the surveyed said their organisation is getting the maximum value from its HR technology.³

Most implementations fail because leaders adopt a ‘tool view’ of technology — they see AI as a faster way to screen resumes, a smarter way to predict turnover. While this view is essential, it is also incomplete as it ignores two fundamental realities. First, employees and candidates are not passive recipients of technology, and second, technology operates within a complex world. Focusing only on the tool will lead to low adoption, employee resistance and ethical blind spots.

Three-lens solution

So how can a company ensure that it can integrate technology with human operations to get the best value? Companies can leverage insights from research by Kim et al. which analyses a new system through three complementary lenses:

Tool view: A perspective that focuses only on technology’s functional efficiency and ROI. This view is the foundational lens that sees technology as a productivity enhancer, which can increase productivity, reduce costs and speed up the processes. Most organisations will evaluate technology based on financial benefits and measurable outcomes. An example of a company utilising a tool view is IBM, which is actively using AI to automate and replace certain back-office HR functions. The company has reported that 94% of routine HR tasks are now handled by AI.⁴ But such a view often doesn’t take into account employee attitudes, organisational culture and social impact.
Proxy view: A perspective that looks at technology as an agent of management, focusing on how employees perceive and respond to it. This ‘human- centric’ view recognises that employees perceive technology such as monitoring software as a ‘proxy’ for management’s intentions, which influences their attitudes and performance.⁵ This view looks at how employees feel about the technology. The employees may adapt to or resist the technology, depending on how they perceive it. This view can affect employee motivation, job satisfaction and trust in management, as workers may question if technology-based decisions (such as AI-enabled performance ratings) are fair and unbiased. A major challenge of this view is ‘productivity paranoia’.⁶ HR leaders fear a loss of productivity they cannot see and implement monitoring tools but employees may feel distrusted and respond with ‘productivity theater’ (endless meetings and visible busywork), which can crush morale and authentic engagement.
Ensemble view: A strategic perspective that considers technology within the broader context of social norms, ethical risks and legal regulations. This view understands that technology is part of a complex ‘ensemble’ of social norms and legal regulations. The goal is not just efficiency but also sustainable, responsible and socially acceptable use of technology. The New York City AI Bias Law (Local Law 144) is a powerful example of this ‘ensemble’ in action.⁷ As of 2023, the law made it illegal for companies to use automated employment decision tools for hiring or promotion of NYC-based candidates without first conducting a rigorous, independent bias audit.

So, what would happen if a company ignores the three- lens solution? A classic tale is Amazon’s scrapped AI recruiting tool.⁸ The AI-based recruitment tool was developed with a ‘tool view’ trained on a decade of hiring data, focused on efficiency, automation and return on investment.

Human Resources

The three-lens solution

Tool view: A perspective that focuses only on technology’s functional efficiency and ROI
Proxy view: A perspective that focuses on how employees perceive, react to and trust technology
Ensemble view: A perspective that looks at technology within the social, ethical and legal contexts

But they did not consider the ‘proxy view’ and ‘ensemble view’, which would account for employee perceptions, ethical concerns, legal risks and social expectations. They also failed to consider that their historical data was heavily biased towards men. The AI learned to penalise resumes containing words like ‘women’, effectively teaching itself that male candidates were preferable. The project was a failure because it was strategically and ethically blind.

The table below synthesises the three lenses into a set of actionable questions and verifiable, real-world examples drawn exclusively from the provided sources.

Table 1: Manager’s diagnostic tool for responsible HR innovation

Lens

Core concept (managerial focus)

Real-world examples: Positive (+) and negative (–)

Key questions to ask

Tool view

Views technology as a neutral instrument for efficiency and productivity
Focus on functionality, features and ROI

+ IBM uses AI to automate and replace certain back-office HR functions to boost efficiency

– 76% of HR tech fails to deliver business value, implying many tools are ineffective

What is the business case?
How much time and money will this save?
Does it perform its core function reliably?

Proxy view

Recognises technology as a ‘proxy’ for management’s intentions
Focus on trust, perception and fairness

+ (Implied) A company co- designing a new monitoring system with employees to build

– Productivity paranoia, where monitoring tools signal distrust and harm morale

How will employees perceive this tool?
How do we build trust in the system’s outputs?
Is it seen as support or surveillance?

Ensemble view

Understands technology exists within a complex social and legal ‘ensemble’
Focus on societal bias, ethics and compliance

+ (Implied) A company proactively conducting bias audits to comply with NYC’s AI Law

– Amazon’s AI recruiting tool learning historical gender bias from past company data

What societal biases could this algorithm amplify?
How do new regulations affect us?
Does this align with our public values?

Source: Created by the author based on the research by Sunghoon Kim, Violetta Khoreva, and Vlad Vaiman. “Strategic Human Resource Management in the Era of Algorithmic Technologies: Key Insights and Future Research Agenda.” Human Resource Management 64, no. 2 (2024): 447–64. https://doi.org/10.1002/hrm.22268, and the examples that are cited in the preceding section

Costs of balanced approach

The framework outlined by Kim and team for integrating AI tools into HR systems works for high-stakes systems that can affect careers, such as hiring and promotion. But there are still some concerns over how companies analyse and integrate AI tools. Some of the most common reasons that the integration may fail are:

Training data that reflects inequalities and biases.
Lack of transparency on how the tool is used, making decisions hard to justify.
No proper checks on the use of AI tools and over- reliance on automation.
No proper interpretation or integration of the output of the AI tools.

Hence, the company needs to invest in robust systems to ensure AI tools can be integrated smoothly. In terms of cost to the company, investing in the framework will require investment in robust AI systems, third-party bias audits, and specialised legal counsel. But the more significant cost involves a deep cultural commitment to ‘human sustainability’ and balancing technology with ethical considerations to build trust.⁹

Here are some steps that can help to better integrate AI with the company’s HR workflow by using the three-lens solution:

AI tools as strategic enablers: AI, machine learning and related algorithmic systems are reshaping HR delivery processes (e.g. recruitment and performance management). But they should be used as enablers that can provide quality data to help managers make strategic decisions.
Efficiency and human oversight: While tools can automate and optimise routine HR tasks, managers should ensure human oversight, especially in decisions affecting employee well-being.
Manage employee perceptions: The organisation should provide clarity about the purpose of AI tools and how they will be implemented. They should address employee concerns about fairness, privacy and perceived loss of control.
Invest in capability building: Companies should start developing internal expertise by training HR staff and line managers in data literacy, ethical use of technology and understanding limitations.
Align with organisational context: HR tools should be tailored to the broader organisational culture and strategic goals.
Monitor outcomes: Regular monitoring of AI tools’ performance and their impact on employees’ well- being is critical. The company should establish guidelines that address gains and also employee trust and engagement.

AI at work

Use of technology for faster HR processing is unavoidable. But the greatest risk to its smooth functioning is a myopic ‘tool view’ that ignores the people and boundaries. The three-lens framework provides the essential guide. As a way to ensure smooth integration of AI tools in HR processes and the company workflow, ask these three questions:

What is the real business problem this solves?
How can we manage employee reactions? How can we build trust?
What is the worst case if the approach goes wrong?

While the three-lens solution provides some answers, leaders must develop solutions that align with their parameters to harness AI tools responsibly, turning a pitfall into a sustainable, strategic advantage.

AI at work

About the author

Saurabh Mehta

Saurabh Mehta is Research Scholar at SPJIMR.

You can reach out to him at fpm25.saurabh@spjimr.org

References

Sunghoon Kim et al., ‘Strategic Human Resource Management in the Era of Algorithmic Technologies: Key Insights and Future Research Agenda’, Human Resource Management 64, no. 2 (2024): 447–64, https://doi.org/10.1002/hrm.22268.
Conn STAMFORD, ‘Gartner Survey Reveals Only 24% of HR Functions Are Maximizing the Business Value from HR Technology’, Gartner, 13 June 2024, https://www.gartner.com/en/newsroom/press-releases/2024-06-13-gartner-survey-reveals-only-24-percent-of-hr-functions-are-maximizing-the-business-value-from-hr-technology.
Conn STAMFORD, ‘Gartner Survey Reveals Only 24% of HR Functions Are Maximizing the Business Value from HR Technology’, 13 June 2024.

Chris Westfall, ‘IBM Replaces Hundreds With AI As HR, L&D Leaders Rethink Roles’, Careers, Forbes, 27 May 2025, https://www.forbes.com/sites/chriswestfall/2025/05/27/ibm-replaces-hundreds-with-ai-as-hr-ld-leaders-rethink-roles/.

Marc King, In Employees We [Must] Trust: Using Employee Monitoring Software for Good and Not Evil – International Association for Human Resource Information Management, 2 June 2021, https://www.ihrim.org/2021/06/in-employees-we-must-trust-using-employee-monitoring-software-for-good-and-not-evil-by-michael-m-moon-phd/.

Karen Grace Larsen, What Is Productivity Paranoia G How Is It Hurting Your Business?, 15 March 2023, https://www.nexushr.com/what-is-productivity-paranoia.

Perspective, NYC Local Law 144-21 and Algorithmic Bias, Deloitte US, 6 April 2023, https://www.deloitte.com/us/en/services/audit-assurance/articles/nyc-local-law-144-algorithmic-bias.html.

Sreevally Pasumarthy, ‘What Can Amazon’s AI Recruiting Tool Teach Us about Reliance on Automation?’, Leoforce, 12 May 2023, https://leoforce.com/blog/amazon-ai-hiring-bias-case-study/.

Olena Boichenko, Deloitte’s ‘2024 Global Human Capital Trends’ Report Identifies Trust and Human Sustainability as Top Issues, Deloitte Ukraine, 7 February 2024, https://www.deloitte.com/ua/en/about/press-room/human-capital-trends.html.

Show All References

Maximizing disability diversity, language diversity, and productivity: A study in apparel manufacturing, 2023; Zhang, Y. Narayanan, S., Soun, T., Deb, K. and Cole, D.

Source