Enterprises in financial services, retail and consumer packaged goods, healthcare and life sciences, telecommunications and manufacturing are rapidly scaling artificial intelligence from pilots to production, according to survey responses from over 3,200 participants worldwide. Overall, 64% of respondents said their organizations are actively using artificial intelligence in their operations, 28% said they are still in the assessment phase and 8% said they are not using artificial intelligence and have no plans to start. North America leads adoption, with 70% actively using the technology, 27% assessing projects and 3% not using artificial intelligence, while adoption in EMEA stands at 65% and in APAC at 63%, with 15% in APAC saying they are not using artificial intelligence. Larger enterprises with more than 1,000 employees show the broadest deployment, with 76% of respondents reporting active usage and only 2% saying they do not use artificial intelligence, reflecting greater access to capital, infrastructure and specialist talent.
Productivity and efficiency are central to why organizations are investing. The top three artificial intelligence goals are creating operational efficiencies (34%), improving employee productivity (33%) and opening new business opportunities and revenue streams (23%). More than half of respondents (53%) said improved employee productivity was one of the biggest impacts on business operations, while 42% said artificial intelligence created operational efficiencies and 34% said it opened up new business and revenue opportunities. In telecommunications, 99% of respondents said artificial intelligence helped improve employee productivity, with a quarter reporting major or significant improvement. In manufacturing, Siemens and PepsiCo are using high fidelity digital twins and artificial intelligence agents to simulate plant operations, identify up to 90% of potential issues before physical changes, deliver a 20% increase in throughput on initial deployments, achieve nearly 100% design validation and realize 10-15% reductions in capital expenditure.
Revenue growth and cost reduction are also emerging clearly from deployments. Overall, 88% of respondents said artificial intelligence has had an impact on increasing annual revenue, with 30% reporting a significant increase greater than 10%, 33% reporting a 5-10% increase and 25% reporting an increase of less than 5%, while a little over 40% of executives said they saw an annual revenue increase of more than 10%. On the cost side, 87% said artificial intelligence helped reduce annual costs, with 25% citing a decrease greater than 10% and 37% of retail and consumer packaged goods respondents saying costs had been reduced by more than 10%. Nearly all organizations expect to maintain or grow investment, as 86% of respondents said their artificial intelligence budget will increase in 2026, 12% said budgets will stay the same and nearly 40% said budgets will increase by 10% or more, with North American organizations and executives especially likely to raise spending by that margin. Top spending priorities include optimizing artificial intelligence workflows and production cycles (42%), finding additional use cases (31%) and building or providing access to artificial intelligence infrastructure (31%).
New workloads and architectures are reshaping how enterprises apply artificial intelligence. Generative artificial intelligence and data analytics dominate usage, with 62% of respondents citing data analytics among their top workloads and 61% citing generative artificial intelligence, which becomes the leading workload in healthcare and life sciences and telecommunications and is the top workload in North America and EMEA. Agentic artificial intelligence, in the form of autonomous agents that reason, plan and execute complex tasks, is moving from experimentation in 2025 to broader deployment in early 2026, with 44% of companies either deploying or assessing agents last year and adoption highest in telecommunications at 48% and retail and consumer packaged goods at 47%. Open source and open weight models underpin many of these strategies, as 85% of respondents said open source is moderately to extremely important to their organization’s artificial intelligence strategy, including 48% who said it is very to extremely important, and 58% of small companies and 51% of executives emphasizing its high importance for building tailored applications and fine tuning models on proprietary data.
Even as adoption accelerates, structural challenges remain. Nearly a third of organizations are still in pilot and assessment phases, and many are struggling to manage and utilize data effectively, with 48% of respondents citing having sufficient data and other data related issues as their top challenge. Lack of artificial intelligence experts and data scientists needed to turn that data into production systems was the next most prominent concern, cited by 38% of respondents. Quantifying benefits can also be difficult, especially for softer outcomes such as knowledge worker productivity, and 30% of respondents cited lack of clarity on artificial intelligence’s return on investment as one of their top challenges. The survey sample spans APAC (32%), North America (26%), EMEA (21%) and the rest of the world (20%), across company sizes from large enterprises with more than 1,000 employees (39%) to mid sized firms with 100-1,000 employees (27%) and smaller organizations with fewer than 100 employees (34%), and includes C suite and vice presidents (27%), directors and managers (33%) and artificial intelligence practitioners (40%), illustrating how pervasive artificial intelligence has become across roles, regions and scales.
