While enterprises are unlocking benefits from AI, a concerning number are still woefully unprepared to expand projects, new research from Kyndryl warns.
The company’s third annual has warned of a “widening divide” between adoption rates and returns on investment (ROI). Nearly two-thirds (64%) of leaders report positive returns on their investments, marking a 10-point increase compared to 2025.
However, average readiness for external risks fell seven points to just 24%. Martin Schroeter, CEO and chairman of Kyndryl, said this means that organizations’ AI ambitions are “moving faster than the systems needed to support them”.
Many, the report found, are failing to address fundamental aspects of AI adoption such as potential cybersecurity risks and infrastructure-related considerations.
During a press briefing ahead of the report launch, Schroeter attributed much of this to the rapid pace of change within the industry, with enterprises now realizing that AI isn’t simply a plug-and-play technology.
“One of the things that’s becoming increasingly clear: adopting new technology and being ready to put it to work are definitely not the same thing,” he said.
“As the rate and pace of change continues to accelerate, businesses need systems, they need operations, and they need people who can adapt just as quickly.”
Sign up today and you will receive a free copy of our Future Focus 2026 report - the leading resource for IT decision-maker insight on priorities and investment areas in AI, security and more.
AI strategies are being overhauled
According to Kyndryl, organizations are making nominal progress with AI, but a significant portion have updated or overhauled strategies to refine adoption processes.
Nearly half (48%), for example, said they “substantially changed” AI strategies over the last year. Similarly, around a quarter (24%) also saw AI investments fail while 21% canceled a major initiative outright.
There are several factors that have a potential impact on AI readiness, according to Kyndryl, and infrastructure capabilities rank among the leading hurdles.
For the third year running, respondents told the firm that upgrading IT infrastructure was the primary action their organization took to compensate for AI adoption.
Infrastructure readiness has become a recurring pain point for enterprises over the last four years, with research showing that outdated IT systems play a key role in failed adoption projects.
This problem still has an acute impact on firms, according to Kyndryl. Nearly three-quarters (71%) rate their infrastructure as “mostly or fully” modernized. However, two-thirds of CIOs said they believe their infrastructure foundations require a “significant re-architecture before being fully ready for AI at scale”.
Security concerns are also having an impact on broader AI readiness, the study noted. Security concerns topped the list of external risks for the third year running, with just 26% of firms completely prepared for cyber attacks.
“Cybersecurity readiness dropped 11 points this year, and it's also the largest drop we saw,” Schroeter said. “That's the tension that's at the heart of this year's report. AI is creating real value, but AI is raising the bar for readiness, and sustaining and growing that value will depend on the readiness of everything around it.”
“It’s not just technology; it's people, it's process, it's how we do what we do. As AI moves deeper into the enterprise, we see three dimensions of readiness becoming increasingly important,” he added.
“Technology foundations that sit underneath, how the organizations and the people work, and the ability to control these technologies and the environments becomes more complex.”
Changing workforce requirements
Intense workforce challenges are a recurring theme throughout the Kyndryl study, Schroeter told assembled media. The report found that as AI projects have matured from experimentation to “sustained adoption”, organizations are falling behind in terms of workforce changes and talent acquisition.
Fewer than one-third (27%) have formally updated roles and responsibilities around AI, which is impacting workforce alignment on adoption processes. A similar number (29%) have fully implemented change management for AI adoption, according to Kyndryl, again representing only a small proportion.
“Much of the conversation has moved to redesigning work and building new skills, but most organizations have not yet embarked on that kind of transformation work,” he said.
“Bringing people along is as important to making AI work as the technology is itself.”
Notably, organizations that have focused heavily on workforce evolution are typically 28 points more likely to be fully prepared for AI implementation. Those same businesses are eight points more likely to report positive returns on investment.