How AI Is Actually Being Adopted Across Industries (2023–2025)

Executive Summary

Artificial Intelligence (AI) adoption has accelerated dramatically in recent years. Globally, approximately 78% of organizations reported using AI in at least one function in 2024, up from 55% in 2023 [1]. Generative AI usage specifically increased from 65% to 71% between the first and second half of 2024 [1]. In the European Union, 13.5% of firms used AI in 2024 (8% in 2023), with large firms reporting adoption rates as high as 41% [2]. By function, AI is most used in IT, marketing and sales, and service operations, with adoption spreading into software engineering and product development [1]. While many business units now report revenue gains and cost reductions from generative AI, enterprise-level EBIT impact remains limited for most companies [3].

On employment, evidence of widespread, AI-caused job losses is limited so far. Companies more often retrain than lay off, though sectors such as media, financial services, and customer support show targeted reductions [4]. Employers expect future displacement and creation to balance out, with forecasts suggesting approximately 9 million jobs displaced but 11 million created globally in coming years [5]. This raises the question: are today’s affected workers better off reorienting themselves toward AI-capable roles?

What AI Adoption Means

When organizations report “using AI,” this may range from simple features such as spam filters or chatbots to fundamental process redesign, such as AI-assisted coding, automated supply chain planning, or AI-enabled customer service. McKinsey, for example, counts any organization with AI in at least one function as an adopter, which blends light experimentation with deep deployments [1].

Adoption Levels: Two Years Ago vs. Today

Globally, 55% of organizations used AI in at least one function in 2023. By mid-2024, that figure had climbed to 72%, and by late 2024 to 78%. Generative AI adoption rose from 65% to 71% in the same period [1].

In the European Union, 8% of firms reported AI use in 2023, increasing to 13.5% in 2024. Adoption differs by company size: 11.2% of small firms, 21% of medium-sized firms, and 41.2% of large firms used AI. The Information and Communication sector leads adoption, with Nordic and Benelux countries among the highest adopters [2].

In the United States, the Census Bureau surveys estimate adoption rose from 3.7% to 5.4% of firms between late 2023 and mid-2024, with expectations of 6.6% by late 2024. This narrower definition (AI “to produce goods or services”) undercounts front-office uses such as marketing and customer service [6].

Where AI Is Used Most

AI is most heavily used in IT, marketing and sales, and service operations, with increasing adoption in software engineering and product development [1].

By industry:

Financial Services: Strong adoption in customer operations, KYC automation, and fraud detection; many banks emphasize retraining [1][4].
Manufacturing: AI is applied in quality inspection, predictive maintenance, and design, though scaling remains uneven [7].
Retail & E-commerce: Applications include personalized recommendations, dynamic pricing, and AI chatbots for customer service [8].
Media & Telecom: Generative AI supports content creation, metadata generation, and customer service, while news publishers face restructuring as AI alters referral dynamics [9].
Healthcare: Adoption in imaging, diagnostics, and administrative workflows, though constrained by regulatory requirements [1].
Aerospace & Defense: Strong investment intentions, with 87% of firms planning to increase AI investment in 2025 [10].
Public Sector: Early use cases in document processing and service portals, though large investments have yet to translate into broad job creation [11].

ROI and Business Value

AI has delivered unit-level value: revenue gains and cost reductions are increasingly reported where it is deployed [3]. For example, conversational AI in customer service can reduce cost per contact by about 23.5% and increase revenue by around 4% [8]. However, enterprise-level impact remains immature, as fewer than one-third of organizations follow best practices such as KPI tracking and clear road maps [3]. Scaling AI effectively requires organizational change and governance, not just technology deployment.

Employment Impact

Evidence of large-scale AI-driven job losses is limited so far. In 2025, the New York Federal Reserve reported that 40% of service firms and 26% of manufacturers in its district had adopted AI, yet only 1% of service firms cited AI-related layoffs, and none in manufacturing. Most opted to retrain or slow hiring instead [4].

In the United States, Challenger, Gray & Christmas reports show that job cuts explicitly citing AI remain a small share of overall layoffs. In July 2025, about 10,000 cuts were attributed to AI, with “technological updates” including AI responsible for roughly 20,000 cuts year-to-date [12]. Media companies have announced editorial role reductions tied to AI, and some fintech firms have publicized replacing parts of customer service with AI, though reversals highlight quality concerns [9].

Forecasts anticipate greater change. The World Economic Forum projects 9 million jobs displaced and 11 million created globally in the coming years, with around 40% of employers expecting workforce reductions in roles subject to automation [5]. The IMF estimates that 40% of jobs worldwide will be affected by AI, either replaced or complemented, depending on policy and reskilling [13].

Industry-Level Labor Effects

The most exposed roles are in customer service, content production, and certain back-office functions. These roles are typically entry-level or junior positions. Media and publishing have seen newsroom reductions, while financial operations and customer support have experienced restructuring. In software engineering, junior hiring appears to be slowing, with firms preferring to upskill existing staff [9][12].

Expert Perspectives

McKinsey notes that fewer than one-third of firms follow most adoption and scaling practices, explaining why enterprise-wide EBIT impact remains rare [3]. IBM’s CEO has predicted AI usage will surge as costs fall, especially among smaller firms [14]. The IMF emphasizes that whether AI complements or replaces jobs depends heavily on national policies and training [13]. Accenture finds that while many executives believe AI will transform their businesses, only 13% report sustained returns from scaling efforts [15].

Conclusion

Adoption of AI is expanding rapidly across industries, reshaping workflows and selectively impacting employment. The greatest effects so far are concentrated in specific tasks and entry-level roles rather than across entire labor markets. Yet forecasts strongly suggest deeper transformations ahead. This raises the central question: given the accelerating pace of adoption, should workers in the most affected categories pivot into AI-capable roles, learning to design, supervise, and integrate AI into workflows now—before hiring patterns and career paths change even further?