>>The impact of AI on Businesses & the Private Sector (Αγγλικά)
by Konstantinos Sioumalas-Christodoulou
AI’s incursion into the private sector is often framed as a sudden rupture. Yet, AI is better understood as the latest surge within a decades-long information and communications technology (ICT) revolution, one that has been mechanising mental rather than manual work (Perez, 2024).
This historical framing matters: it reminds us that the real question is not whether AI will transform business, but who will benefit, under what institutional arrangements, and with what distributional consequences.
In practice, AI is already reshaping firms along three main axes:
- Automation of routine cognitive tasks,
- Augmentation of knowledge work, and
- Data-driven reconfiguration of products and business models.
Field evidence shows that generative AI tools can boost frontline productivity: customer-support agents given an AI assistant resolve around 15% more issues per hour on average, with the largest gains for less-experienced workers (Brynjolfsson et al., 2025). At the firm level, surveys across G7 countries indicate that adopters are using AI for logistics optimisation, predictive maintenance, customer analytics, and R&D, with expectations of faster innovation and greater process efficiency (OECD/BCG/INSEAD, 2025).
However, the current trajectory is far from an inclusive “golden age”. First, adoption is uneven: large, digitally mature firms pull ahead, while smaller manufacturers and service firms face steep barriers; uncertain returns on investment, difficulties in accessing skills and data, and dependence on a small pool of vendors (OECD/BCG/INSEAD, 2025). Second, a historical lens suggests we are behaving as if AI were a brand-new rupture, fuelling financial hype, platform concentration and short-term speculation, rather than using it to complete the unfinished project of the ICT revolution: widely shared productivity and green restructuring (Perez, 2024). That explains why this AI moment so readily takes the form of a bubble: capital flows toward fast monetization and narrative momentum, not toward the slow diffusion of productivity (Floridi, 2024). Third, distributional risks are real. Task-based modelling suggests AI could compress wage gaps by automating high-wage tasks, yet simultaneously amplify wealth inequality as those same workers disproportionately own the capital and equity that reap AI-driven returns (Rockall et al., 2025).
Taken together, these patterns suggest that AI's impact on business is not just a technology story but a political-economy choice. If left to narrow financial logics, AI will mainly be deployed for cost-cutting, worker surveillance and market power, entrenching existing asymmetries between large platforms and the rest of the economy. If steered differently, it could diffuse sophisticated capabilities into SMEs, enable new forms of worker learning, and be harnessed for decarbonization and social infrastructure.
A different trajectory requires coordinated action. Industrial and competition policy should link AI incentives such as tax breaks, public contracts and access to high-quality public data (UNCTD, 2025), directly to clear public goals like climate transition, stronger health systems, and education, as well as to broader ESG commitments. Access to these incentives should depend on whether projects deliver concrete improvements in workers’ jobs and livelihoods, not just higher profits. This, in turn, requires robust, disaggregated metrics that track not only productivity and profitability, but also job quality, wage distribution, skill development and environmental impact, enabling policymakers to monitor who actually benefits from AI deployment and to recalibrate incentives when systems deepen rather than reduce inequality.
Governments should scale “AI extension services”: public or public-private institutions that help smaller firms identify use cases, experiment safely, and build in-house capabilities, a role early evidence shows to be effective but underprovided (OECD/BCG/INSEAD, 2025). At the corporate level, boards need to treat AI strategy as a socio-technical issue, embedding worker participation, transparency over data use, and long-term capability building rather than chasing short-term head-count reductions. Finally, macro-level distributional risks call for complementary policies - stronger taxation of economic rents, broadened capital ownership, and robust social protection to ensure that the gains from AI adoption are not captured solely by a thin stratum of firms and asset holders (Rockall et al., 2025).
REFERENCES
Brynjolfsson, E., Li, D., & Raymond, L. (2025). Generative AI at work. The Quarterly Journal of Economics, 140(2), 889–942. https://doi.org/10.1093/qje/qjae044
Floridi, L. (2024). Why the AI hype is another tech bubble. Philosophy & Technology, 37(4), 128.
OECD/BCG/INSEAD. (2025). The adoption of artificial intelligence in firms: New evidence for policymaking. OECD Publishing. https://doi.org/10.1787/f9ef33c3-en
UNCTD (2025). United Nations Conference on Trade and Development. Technology and innovation report 2025: Inclusive artificial intelligence for development. Αvailable at: https://unctad.org/system/files/official-document/tir2025overview_en.pdf
Perez, C. (2024). What is AI’s place in history? Project Syndicate. Available at: https://www.project-syndicate.org/magazine/ai-is-part-of-larger-technological-revolution-by-carlota-perez-1-2024-03
Rockall, E. J., Mendes Tavares, M., & Pizzinelli, C. (2025). AI adoption and inequality (IMF Working Paper No. 2025/068). International Monetary Fund. https://doi.org/10.5089/9798229006828.001