When it comes to AI and the mainframe, customers are moving from experimental to operational approaches and feeling comfortable enough to use AI to provide system insights and recommend actions, a new study finds.
Mainframe professionals are increasingly adopting generative AI tools to recommend actions, according to the 2026 BMC Mainframe Survey, and implementing AI technologies remains a top priority for 45% of respondents. BMC surveyed more than 1,300 mainframe practitioners and decision makers globally for the report.
“We are seeing a significant shift as organizations have gone from asking how they can use AI to asking where they can trust it, how it can be governed, and where it delivers measurable value. It’s becoming clear that the path to greater AI autonomy on the mainframe will be earned through trust, with the option of humans remaining in the loop to oversee and implement AI recommendations,” said John McKenny, senior vice president and general manager of Intelligent Z Optimization and Transformation at BMC, in a statement.
After much discussion, planning and investment, the mainframe community is moving from “AI enthusiasm to pragmatic adoption,” according to the report:
“AI has moved from experimentation to strategic planning, with mainframe organizations seeming to take a more pragmatic approach,” the report states. “That pragmatism is visible in the kinds of responsibilities organizations are currently willing to give AI. After the initial excitement and hype concerning AI capabilities, mainframe executives seem to be adopting a cautious realism, turning to the technology to act as an advisor, but not an executor. In short, AI hasn’t gained the full trust of the mainframe world, prompting a ‘human-in-the-loop’ approach.
AI is becoming an increasingly important tool in mainframe transformation, and as expectations mature, “AI use has shifted from experimental to operational; organizations have gone from asking how they can use AI to asking where they can trust it, how it can be governed, and where it delivers measurable value,” the report states.
When it comes to applying AI on the mainframe, the top four concerns when implementing AI-driven solutions are: high implementation costs (cited by 41% of respondents); security and privacy (39%); data integration issues (37%); and regulatory/compliance concerns (22%).
Digital certificate management is one area where many organizations still have an opportunity to improve, the study notes: “As the integration of AI-based tools and use of AI agents increases, so will the need for more, and more frequently renewed, digital licenses, likely reaching a volume to which manual management efforts cannot scale and a complexity that exposes the limitations of home-grown automation solutions.”
“At the same time, connections with AI-based applications will continue to increase as the technology is further integrated with the mainframe,” the report states. “While ensuring secured connections for both humans and applications is a critical piece of the mainframe security puzzle, the 2026 BMC Mainframe Survey shows that the majority of respondents are using either in-house automated solutions to manage their organization’s digital certificates or are managing certificates manually.”
Some other interesting findings in the study include: