By David De Cremer, Mark Esposito, and Emilios Galariotis
Artificial intelligence (AI) continues to reshape the workplace, with the dominant narrative emphasizing gains in efficiency and productivity. For example, a Morgan Stanley survey (2026) among 935 corporate executives in the U.S., Germany, Japan and Australia working in five sectors identified as most exposed to AI adoption (consumer staples distribution & retail; real estate management and development; transportation; healthcare equipment and services; and automobiles and components) revealed that these companies reported an 11.5% increase in net productivity because of the use of AI.
While AI undeniably automates routine tasks, reduces costs, and theoretically frees up employee time for higher-value work, emerging evidence suggests a critical flaw in current implementation paradigms—one that risks harming organizations and their workers if left unaddressed. The first concern is that the urge to automate as quickly as possible to obtain productivity gains at a lower cost is a finite strategy. Indeed, each dollar invested in automation will eventually generate smaller incremental returns. Therefore, the motivation to automate for reducing labor costs, and thus replace humans with AI, will ultimately leave organizations more vulnerable at the long term. In fact, a first sight automation seems the rational strategy to follow as it will leave an organization doing the same operational things as before but at a higher speed. However, over time, the fast automation strategy will lose the innovation potential that humans bring. Consequently, the competitive edge of the company will be eroded and therefore fail to achieve the goal of maximizing firm value. For obvious reasons, employees fear such scenario and resist. In addition to resisting the idea of AI and automation efforts by the organization, employees will also suffer at a behavioral level where adapting to AI and keeping up with the pace of change will come with psychological constraints and pressures.
The Efficiency Paradox: When Productivity Gains Backfire
The prevailing discourse frames AI as a tool for maximizing output, assuming time saved through automation will translate into reduced workloads. However, reality reveals a different pattern: rather than decreasing work hours, organizations often reinvest efficiency gains into expanded expectations (Kim & Lee, 2024). They will expect, with the help of AI, to do more of the same. If this strategy is adopted, then AI is not used to liberate people and unleash the creativity potential of employees. Rather, AI will then be used to address urgent operational tasks by optimizing the existing human capabilities to do more of those tasks at higher speed. Under those circumstances, AI is not employed to create room and time for employees to foster their creativity to approach challenges and opportunities in more innovative ways; which is needed for a company to grow and stay competitive. As such, a status quo situation emerges where companies become more productive in what they are doing today, but because they are not investing in humans (rather treating humans as accessories to AI; a kind of adapted robots), they are not able to re-invent themselves and engage in organization transformation and future development. It is at this point that employees find themselves pressured to accomplish more in the same timeframe, leading to stress, burnout, and diminished wellbeing.
This phenomenon—what we term the efficiency trap—manifests across industries. Research by Brynjolfsson and McAfee (2014) demonstrates that while technology drives productivity growth, it frequently results in work intensification rather than workload reduction. A 2023 Upwork Research Institute survey found that 77% of employees report AI has increased their workloads, and a University of Lausanne study revealed that over half of time saved through AI is lost to cognitive recovery needs—workers require breaks and social connection to sustain performance (De Cremer & Koopman, 2024). Without intervention, AI risks fueling a culture of burnout rather than liberation.
The Hidden Costs of Narrow Efficiency Metrics
A singular focus on output maximization carries severe consequences for employee health and organizational resilience. Chronic overwork erodes mental and physical wellbeing (WHO, 2021) while stifling creativity and engagement—key drivers of innovation (OECD, 2023). The economic implications are equally concerning and collectively affect the value of the firm. And, if this scenario unfolds, the negative externalities for society and its organizations are multiple, including:
- Short-term productivity gains may give way to long-term attrition and disengagement. A study by Deloitte found that companies with high burnout rates see turnover costs that offset AI-driven efficiency benefits (Hampson, Cruz, Katzer, and Matyaszek, 2024).
- Job insecurity and substitution fears arise when AI is framed as a cost-cutting tool, undermining motivation and performance (Frey & Osborne, 2017).
- A continuous communication strategy of companies adopting AI where the importance and necessity of AI for the organization is stressed may backfire and result in AI fatigue – a phenomenon that ultimately leads to employees being less willing to learn about AI and experiment with it; all of this to the detriment of any successful AI adoption project (De Cremer, Chan, & Kim, 2025).
This reality creates a paradox: organizations leveraging AI for efficiency may inadvertently sabotage the very human capital required to sustain growth.
A Human-Centered Alternative: Augmentation Over Automation
To avoid this trap, we must redefine success in the AI era—moving beyond productivity metrics to prioritize human wellbeing, autonomy, and meaningful work. The European Commission’s Ethics Guidelines for Trustworthy AI (2019) provides a blueprint, emphasizing human dignity, agency, and flourishing as core design principles to assess the successfulness of an AI integration effort.
Three Pillars of Responsible AI Integration
Integrating AI in the workplace needs to preserve human workers’ wellbeing and innovative capacity (Kim and Lee, 2024) and therefore organisations must encourage employees to use AI as a tool serving personal development that can feed into their intrinsic motivation. In other words, creating work conditions under which the use of AI will facilitate employees to become better at what they are already good at and therefore stay competitive and open to learning and innovating. In fact, as noted by the McKinsey Global Institute, organisations that focus on employee wellbeing and development are more likely to see sustained long-term productivity gains as employees become more motivated, engaged, and innovative (Manyika, J., Chui, M., Miremadi, M., Bughin, J., George, K., Willmott, P., and Dewhurst, M., 2018). In line with such an approach, we argue that when AI is adopted within the realm of employee’s jobs, organizations should safeguard this kind of intrinsic motivation by satisfying and empowering the following needs:
1. Autonomy Preservation
- Employees must retain meaningful control over AI-assisted workflows.
- Transparent communication—positioning AI as a support toolrather than a replacement—mitigates stress and sustains engagement (Kreacic et al., 2024).
2. Belongingness Reinforcement
- AI-driven workflows risk isolating employees in cognitively demanding, socially detached tasks.
- Counteract this with structured social breaks, collaborative AI feedback sessions, and “human-first” workspaces(McKinsey, 2023).
3. Competence Development
- Frame AI as an augmentation tool, freeing employees to upskill rather than obsolesce.
- Invest in continuous learning programs aligned with evolving job demands (Manyika et al., 2018).
Policy Imperatives for a Balanced Future
Policymakers must ensure AI adoption benefits both businesses and workers. UNESCO’s 2021 guidelines call for:
- Worker protections against surveillance and overwork.
- Reskilling initiatives to future-proof labor markets.
- Ethical AI standards prioritizing human rights over pure efficiency.
Conclusion: AI as a Catalyst for Human Potential
The true promise of AI lies not in squeezing more output from workers but in redesigning work for human thriving. By balancing the long term interests of the company with efficiency as well as with wellbeing, organizations can unlock sustainable innovation—where technology elevates rather than exploits human potential. The path forward demands bold leadership, ethical frameworks, and a commitment to measuring success in human terms as much as economic ones (De Cremer, 2024).
This conclusion raises an important and even broader challenge for business leaders today, which is the need to keep learning and reflecting on the conditions that AI creates for humans. If AI pushes organizations and its leaders to mainly pursue immediate efficiency while ignoring its longer-term human and organizational consequences, the challenge for AI to further humanity will not lie so much in the intelligence of the technology, but rather in the myopia of those who govern it. The future of work will therefore depend not only on how intelligent our machines become, but on whether human judgment becomes wise enough to use them well to make humans and organizations grow (which we call “humanAIzation”; De Cremer & Esposito, 2025).
About the Authors
David De Cremer is a chaired research professor at King Fahd University of Petroleum and Minerals (KFUPM, Saudi Arabia, a research affiliate at MIT (Center for Collective Intelligence) and Yale Law School (Justicecollaboratory) and an honorary fellow at St. Edmunds College, University of Cambridge. He is also a founding member of EY’s Global AI advisory board. Before moving to KFUPM, he was the Dunton Family Dean and professor of management and technology at D’Amore-McKim School of Business, Northeastern University (Boston), a Provost’s chair and professor in management and organizations at NUS Business School, National University of Singapore, and the KPMG endowed chaired professor in management studies at Cambridge University. He is the founder of the Center on AI Technology for Humankind in Singapore and named one of the World’s top 30 management gurus and speakers by the organization GlobalGurus, one of the “Thinkers50 list of 30 next generation business thinkers” and continuously included in the World Top 2% of scientists. He is a best-selling author with his new book “The AI-savvy leader: 9 ways to take back control and make AI work” (published by Harvard Business Review Press) which was the winner of the Outstanding Literature Award (category leadership) in the US and included in the Forbes list of top 10 AI books in 2024.
Mark Esposito is teaching professor at D’Amore-McKim School of Business, Northeastern University and Faculty Affiliate at Harvard’s Center for International Development at the Kennedy School. He serves as Chief Economist of micro1, a Silicon Valley AI lab, and is a Senior Fellow at the Institute for Peace and Diplomacy. He is a member of the World Economic Forum’s Global AI Alliance and Fostering Converging Technologies group, and has been a Professorial Fellow at the Mohammed Bin Rashid School of Government in Dubai since 2017. As an entrepreneur, he co-founded Nexus FrontierTech and TheChart Thinktank.
Emilios Galariotis obtained his PhD from Durham University Business School (UK) and his HDR from the University of Nantes in France. Currently, he is a Full Professor of Finance at KFUPM in Saudi Arabia. Prior to moving into academia he was in the banking industry and, for more than twenty years, in triple accredited, Russel Group and FT-Ranked Business Schools in the UK and France, and as a distinguished research chaired professor at KIMEP University in Kazakhstan. He is also a member of the management board of the Greek tax and customs authority IAPR (Independent Authority for Public Revenue) of the Hellenic Republic. He has authored and co-authored eighteen book-chapters including one in the Wiley Encyclopaedia of Management, and co-authored/edited five books including one in the Frank J. Fabozzi Series. His research has been published in the Journal of Corporate Finance, British Journal of Management, Journal of Business Ethics, European Journal of Operational Research, The Journal of Banking and Finance, Technological Forecasting and Social Change, Annals of Operations Research, Journal of Development Studies, Journal of Economic Behavior and Organisation, etc.
References
Brynjolfsson, E., and McAfee, A. (2014) The second machine age: Work, progress, and prosperity in a time of brilliant technologies. New York: W.W. Norton & Company.
De Cremer, D. (2024). The AI-savvy leader: 9 ways to take back control and make AI work. Harvard Business Review Press.
De Cremer, D. (2026). When developing an AI strategy, beware the urgency trap. Harvard Business Review. Available at: When Developing an AI Strategy, Beware the Urgency Trap
De Cremer & Esposito, M. (2025). HumanAIzation: The art of embracing AI to become more human. California Management Review. Available at: HumanAIzation: The Art of Embracing AI to Become More Human | California Management Review
De Cremer, D., & Koopman, J.(2024). Using AI at work makes us lonelier and less healthy. Harvard Business Review. June 24.
European Commission. (2019) Ethics guidelines for trustworthy AI. Available at: https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai (Accessed 5 December 2024).
Frey, C. B. and Osborne, M. A. (2017) ‘The future of employment: How susceptible are jobs to computerisation?’, Technological Forecasting and Social Change, 114, pp. 254-280. https://doi.org/10.1080/1755182X.2010.523145
Hampson, E., Cruz, M. J., Katzer, V., and Matyaszek, A. (2024) Mental health and employers. Available at: https://www2.deloitte.com/content/dam/Deloitte/uk/Documents/consultancy/deloitte-uk-mental-health-report-2024-final.pdf (Accessed 5 December 2024).
Kim, B.-J., and Lee, J. (2024) ‘The mental health implications of artificial intelligence adoption: the crucial role of self-efficacy’, Humanities and Social Sciences Communications, 11, Article 1561. https://doi.org/10.1057/s41599-024-04018-w
Lemos, S.I.C., Ferreira, F.A.F., Zopounidis, C., Galariotis, E., and Neuza, C. (2025). Artificial intelligence and change management in small and medium-sized enterprises: an analysis of dynamics within adaptation initiatives. Annals of Operations Reseearch 353, 197–223 (2025). https://doi.org/10.1007/s10479-022-05159-4
Kreacic, A., Jesuthasan, R., and Romeo, J. (2024) 3 ways companies can mitigate the risk of AI in the workplace. Available at: https://www.weforum.org/stories/2024/01/how-companies-can-mitigate-the-risk-of-ai-in-the-workplace/ (Accessed 5 December 2024)
Manyika, J., Chui, M., Miremadi, M., Bughin, J., George, K., Willmott, P., and Dewhurst, M. (2018) Harnessing automation for a future that works. Available at: https://www.mckinsey.com/featured-insights/digital-disruption/harnessing-automation-for-a-future-that-works (Accessed: 5 December 2024).
Morgan Stanley (2026). AI’s impact accelerates. Available at: AI Adoption Surges Driving Productivity Gains and Job Shifts | Morgan Stanley
Organisation for Economic Co-operation and Development. (2023) The impact of AI on the workplace: Evidence from OECD case studies of AI implementation (OECD Social, Employment and Migration Working Papers No. 289). Available at: https://www.oecd-ilibrary.org/docserver/2247ce58-en.pdf?expires=1733494683&id=id&accname=guest&checksum=8FC3A8BC73A6117933200D4406FE6CB7 (Accessed 5 December 2024).
United Nations Educational, Scientific and Cultural Organization. (2021) Recommendation on the Ethics of Artificial Intelligence. Available at: https://unesdoc.unesco.org/ark:/48223/pf0000381133/PDF/381133eng.pdf.multi.page=3 (Accessed 5 December 2024).
United Nations Educational, Scientific and Cultural Organization. (2022) Approval of the Draft Comprehensive Strategy for the MOST Programme, 2022-2029 (MOST/IGC/2022/008). Available at: https://unesdoc.unesco.org/ark:/48223/pf0000381417.locale=en (Accessed 5 December 2024).
World Health Organization (2021) Long working hours increasing deaths from heart disease and stroke: WHO, ILO. Available at: https://www.who.int/news/item/17-05-2021-long-working-hours-increasing-deaths-from-heart-disease-and-stroke-who-ilo (Accessed: 5 December 2024).
