While finance leaders address market volatility and economic uncertainty, investment in AI continues at a heated pace. Under chaos, adoption accelerates. Progress of the use case. Economic benefits arise slowly.
Stanford University's Human-Centered AI Institute recently released its 2025 AI Index Report and Survey.
The annual report summarizes information on AI model development, business use, ethical issues, government policy changes, and the expected economic impact of AI.
We have drawn out some key facts, trends and takeout for our finance leaders. They are key to developing AI use cases within the organization and building pathways for IT.
Big dollar. In 2024, US private AI investments rose to $109.1 billion. Generation AI has had strong momentum, winning $33.9 billion worldwide, up nearly 19% since 2023. AI business use is also accelerating. 78% of organizations reporting using AI in 2024 increased from 55% the previous year.
Growth of AI models. The US remains a major source of AI, with the agency producing 40 notable AI models last year. And almost 90% came from the industry. “AI models are becoming ever bigger and bigger, more computationally demanding and more energy-intensive,” says Stanford Report. The model scale is growing rapidly. “Training doubles every five months, and datasets doubling the power every eight times.”
Business development. The most common applications are content support for marketing strategies (27%) followed by knowledge management (19%), personalization (19%) and design development (14%). Most of the major reported use cases are within the marketing and sales functions. Survey respondents report using generated AI in at least one business feature more than doubled from 33% in 2023 to 71% in 2024.
Faulty reasoning. “Complete reasoning remains problematic” for large-scale language models, the report states. These “systems cannot use logical inferences such as arithmetic and planning to ensure that problems where empirically correct solutions are found using logical inferences, particularly when they are larger than trained,” reports show that reliability affects the “fitness in high-risk applications” of these systems.
Economic impact. Approximately half (49%) of respondents whose organizations use analytics AI in service operations report cost savings followed by supply chain management (43%) and software engineering (41%). Most of them report cost savings of less than 10%, according to the Stanford Report. Regarding revenue, “71% of respondents using AI in marketing and sales report revenue, 63% in supply chain management and 57% in service operations.” The most common level of revenue was below 5%.
Social emotions. “Although AI is considered a timesaver and entertainment booster, questions remain about its economic impact.” Most surveyed people are very confident in their health and economic benefits. Only 38% believe that AI will improve health. Over a third (36%) believe AI will improve the national economy, with 31% having a positive impact on the job market and 37% will strengthen their jobs.
This article was first published in CFO Leadership's Financial and Accounting Technology Briefing.
