Only a quarter of AI initiatives’ expected return on investment ( ROI ) has been met, according to a recent IBM study, and only 16 % of them have successfully scaled AI across the enterprise despite rapid investment and growing competition pressure.
Despite this, the majority of CEOs however support AI, believing that it is essential for long-term success, with the majority anticipating profitable results by 2027. According to IBM Vice Chairman Gary Cohn, “leaders who don’t use AI and their own information efficiently choose not to engage.”
Executives continue to invest in AI despite receiving subpar results.
According to the 2025 IBM CEO Study, 2, 000 Executives from 24 different locations and 24 sectors responded to the survey, the majority of leaders continue to invest in AI even though only a small percentage of projects achieve the expected returns.
CEOs continue to give priority to AI as a key driver of future growth, with 77 % predicting a positive result in scaled AI growth and expansion by 2027 and 85 % anticipating a positive ROI in scaled efficiency and cost savings.
The implementation of relational AI and AI agents is growing faster.
Additionally, according to the IBM study, 61 % of CEOs are actively adopting and preparing to use AI agents scale, believing that AI is essential to maintaining their position in the market.
This widespread adoption is in line with IBM’s new statement that the “era of AI research is over,” as businesses transition from small-scale tests to incorporating AI agents into routine operations to improve modeling, agility, and decision-making across the organization.
What’s stopping AI from paying off?
stifled technology and information environments
Many businesses still deal with fragmented data and disconnected business functions, which impede the impact of AI. According to a recent study, leading CEOs address this by integrating end-to-end workflows and creating a unified information setting, which is essential for unlocking AI’s complete value across teams.
Privacy and security
As businesses expand the use of artificial intelligence, security and data protection continue to be among the best issues, sharing the concerns of security professionals. According to the IBM survey, one of the best CEO objectives for 2025 is improved security and data privacy.
measuring ROI difficult.
Many leaders challenge to compare and compare AI investments to tangible business outcomes, which another businesses across industries even struggle with. However, some business leaders think that real-time, AI-driven ROI is possible, especially for those who invested early and made the correct use cases.
How to get AI ROI
According to the IBM Study, leaders who are willing to take proper risks and use their data to gain will be the ones who will see AI’s whole payoff. According to IBM, Executives who hold may be choosing not to thrive because they don’t really miss out on productivity increases.
In the end, AI’s benefit will be determined by the smart decisions, not just the strong ones.