For three years, a study by Wharton Human-AI Research and GBK Collective has been tracking how generative AI (Gen AI) is gaining ground in businesses. In 2023, only 37% of executives used Gen AI at least once a week, but today that figure has reached 82%. Some 46% of respondents work with it daily, an increase of 17 percentage points compared with last year. This shift means that the phase of experimentation and cautious testing is over—AI has become part of routine operations at many companies with more than 1,000 employees and revenue exceeding $50 million.
The study, led by Jeremy Korst, Stefano Puntoni, and Prasanna Tambe, included 800 respondents from various industries in the US. The respondents were senior managers from areas such as HR (human resources), IT, legal, marketing, operations, product development, procurement, finance, and general management. The survey was conducted from June 26 to July 11, 2025, and focused on actual AI usage, investments, and impacts.
Most Common Uses of AI
Businesses are focusing primarily on tasks that support employee productivity. Data analysis is the most popular use case, cited by 73% of users, followed by summarizing documents or meetings at 70%, and writing or editing text at 68%. These applications are not only the most common but also the highest-rated in terms of performance. For example, IT departments often use AI to write code, HR uses it for employee recruitment and onboarding, and legal teams use it to generate contracts.
Differences are evident across departments. IT and procurement departments lead in both frequency of use and confidence when working with AI—for example, operations saw a 24-percentage-point increase in those rating themselves as competent or expert. By contrast, marketing, sales, and operations are lagging behind, a trend that has persisted since the first wave of the study in 2023. Large companies with revenue exceeding $2 billion have caught up with smaller businesses, but industries such as technology, telecommunications, banking, finance, and professional services are ahead, while manufacturing and retail remain behind.
Seniority plays a role—vice presidents and higher-level executives are more optimistic and twice as likely to believe that their companies are deploying AI faster than their competitors (56% compared with 28% among middle management).
Investment in AI
Three-quarters of executives report a positive return on investment in AI, or 75%. Measuring ROI (return on investment) has become standard practice—72% of companies track metrics such as profitability, throughput, or workforce productivity. Some 88% of companies are expected to increase their AI budgets over the next 12 months, with 62% planning an increase of 10% or more.
Money is shifting from pilot projects to proven programs. On average, 30% of AI technology budgets goes toward internal research and development, indicating an effort to create proprietary solutions. Vice presidents are more optimistic about ROI (81% see positive results compared with 69% of middle managers). Smaller companies with revenue below $250 million report faster results, while the largest enterprises often say it is still too early to assess them.
People as the Key Factor for Success—the Mistake Made by Lagging Companies
Stefano Puntoni, professor of marketing at the Wharton School, says: “The challenge is not replacement but readiness. Companies that invest in training, culture, and safeguards will be the ones that turn everyday AI into a long-term advantage.” The study shows that 89% of leaders see AI as a tool that enhances employees' skills, while 71% acknowledge that it replaces some skills. Nevertheless, 43% warn of the risk of declining proficiency if companies fail to invest in practice.
Lagging companies (the 16% of respondents who use AI less than weekly) cite strict workplace restrictions, low trust, budget pressures, and slow-moving industries as the main obstacles. Leading companies, by contrast, emphasize open access to tools, rapid implementation, and clear rules. Leadership is becoming stronger—67% of companies have a C-suite executive responsible for AI (an increase of 16 percentage points), and 60% have a CAIO (Chief AI Officer) role.
Training, however, is lagging—investment in it fell by 8 percentage points, and confidence in it as a path to fluency declined by 14 percentage points. Recruiting advanced talent is a challenge for 49% of companies. Managers expect an impact on junior roles—17% anticipate fewer interns, but 49% expect more. Middle management prefers an employee-led approach, with greater investment in training (by 12 percentage points) and innovation (by 11 percentage points) compared with vice presidents.
Safeguards are improving—64% of companies have data security policies (an increase of 9 percentage points), and 61% have awareness training programs (an increase of 7 percentage points). AI is also used for risk management purposes, such as fraud detection (62%, an increase of 7 percentage points) and risk management (59%, an increase of 5 percentage points). People remain the key factor—either an obstacle to or an accelerator of AI adoption.
