In 2017, one in five businesses surveyed reported using AI in “at least one business area,” compared to one in two today, according to McKinsey’s annual Global Survey on AI. But that’s not the only thing that’s grown: The average number of AI “capabilities” those businesses are using has also doubled from 1.9 to 3.8, with robotic process automation, computer vision, and natural language text understanding leading the charge.

Business investment in AI has continued uninterrupted despite a general market downturn.

52% of organisations reported investing 5% or more of their digital budgets on the technology, up from 40% of businesses adopting AI in 2017. And 63% anticipate increasing their corporate AI spending until 2025.

According to the report, businesses reported seeing the greatest revenue effects from using AI in marketing and sales, the development of products and services, and strategy and corporate finance, with “the highest cost benefits from AI in supply-chain management,” given the current state of the world’s supply-chains.

According to the survey, most organisations believe that it hasn’t gotten any simpler to fill AI-related tasks this year, suggesting that the talent scarcity for the field is still present. However, there is also evidence to suggest that diversity in AI is still woefully missing.

Last but not least, a sizable portion of AI-using companies can receive a failing grade in the “risk mitigation” area. Despite an increase in AI usage, reported mitigation of any AI-related hazards has not significantly increased since 2019—the year we started collecting this data—despite an increase in AI deployment.