A new report shows that 83% of S&P 500 companies cited AI as a risk in 2025, up from 12% just two years prior, according to The Conference Board.
At the same time, AI expertise remains low: Only 2.7% said they have expertise last year, up from 1.5% in 2021. In contrast, cybersecurity expertise nearly doubled to 27% and tech expertise rose to 51% from 20% over the same 4-year period.
As of December 2025, a quarter of companies said their board has “low or no fluency” in AI, while another quarter said they are “highly fluent” in AI. Half said they have moderate fluency.
To address the AI expertise gap, 26% of executives said they plan to educate the board and upskill the workforce in AI. However, only 21% plan to add AI experts to the board.
The organization based its report on S&P 500 disclosure data as of December 2025 and a survey of 130 executives, mainly from large U.S. publicly held companies.
The biggest AI risk comes in the form of cybersecurity and data breaches, cited by 58% of respondents. A third said privacy and data protection, while 28% said operational and implementation risk and 27% pointed to legal liability and litigation exposure.
As for how AI will affect employees, 75% of executives said AI will disrupt employment and workforce structures “on a large scale” in the next three years. Also, 41% said AI will worsen inequality within countries and among nations.
But AI adoption should raise productivity and efficiency across the economy in the next three years, according to 80% of respondents.
Which net impact of AI will win out – positive or negative? Six out of 10 believe that responsible governance and regulation will determine the outcome.