Security and AI: What Keeps IBM Technologists Up at Night
The 2026 TechChannel Reader Survey shows IBM Z and IBM Power professionals have put AI to work—but they haven't stopped checking its answers
Trust is hard to earn in enterprise IT. TechChannel readers whose companies run IBM Z and IBM Power servers grant it carefully, according to our recent reader survey.
From mid-September to mid-October, we surveyed TechChannel readers. A total of 136 took the time to tell us how they use AI, what worries them and what they want from TechChannel. Their answers paint a picture of a seasoned, influential community.
Who We Heard From
Of the respondents, 81 (60%) use IBM Power regularly and 55 (40%) use IBM Z. Each group answered its own set of questions, which lets us compare the two communities side by side.
Our readers are experienced. More than half of both groups are between 55 and 66. Factor in everyone 45 and older, and that figure climbs to 80% of Z readers and nearly 90% of Power readers. IT staff, systems analysts and programmers, and IT managers make up the largest share of job functions in both groups.
They also work in different kinds of organizations. Nearly half of Z readers (49%) work at companies with 5,000 or more employees, concentrated in IT services, consulting and banking. Power readers are spread more evenly, from small shops to large enterprises, with IT services, banking, manufacturing and wholesale distribution leading the way.
What the two groups share is influence. Seventy-one percent of Z readers and 78% of Power readers have a hand in purchasing decisions, whether they approve budgets, make technical recommendations or shape the conversation. Nearly half of Power readers (49%) make technical recommendations on new hardware and software.
AI Is at Work, but on Probation
AI is no longer an experiment for this audience. Three in four Z readers use it at least weekly for work, including 47% who use it daily. Power readers aren’t far behind: 69% use AI weekly or more, and 38% use it daily. A surprisingly small number–7% of Z readers and 12% of Power readers–say they never use AI for work.
Trust is also a factor. Asked, on a scale of 1 to 5, how much they trust AI output for work-related tasks, both groups landed in the low 4s. Readers rely on AI every day without taking its answers at face value.
One Z reader summed up the mood with a nod to The Hitchhiker’s Guide to the Galaxy, adding that AI is “trust but verify at this stage.” A Power reader worried about users officially reporting “hallucinated AI generated analysis.”
That skepticism isn’t resistance, nor is it surprising. It’s the posture you would expect from people who run systems where a wrong answer can stop production. And it hasn’t dampened their curiosity. When we asked readers which non-IBM technologies interest them most, AI was the top choice for just over half of both groups.
What Keeps Readers Up at Night
We asked an open-ended question: “As a technologist, what keeps you up at night?” We asked AI to group the answers into themes, counting responses that named more than one worry in each category that applied.
Security tops the list. Security and cyber threats drew 32 mentions, more than any other concern. Many answers were a single word—”security,” “hackers”—but ransomware came up repeatedly. Others went deeper. “Am I fully prepared for a cyber attack and can I recover?” one Z reader asked.
For Z readers, AI is the No. 1 worry. AI drew 15 mentions from Z readers, more than security or uptime, compared with nine from Power readers. Their concerns ranged from the broad impact of AI to low-quality output and AI security, including agentic AI as a new kind of “digital worker” to secure. A few saw opportunities instead, such as building generative and agentic AI use cases for the mainframe.
Uptime is never far from mind. Downtime, production stoppages, performance problems and disaster recovery drew 23 mentions across both groups. That’s no surprise from people who run the systems that can’t go down. One Z reader in their 60s said they only lose sleep when something is actively failing, and “cannot wait for 7:30 a.m.”
The workload keeps growing. Power readers in particular described the strain of backlogs, constant change and keeping up with patches, new technology and security. Fifteen Power readers raised workload or the pace of change, putting it on par with uptime.
Shiny Objects and Succession Plans
Some of the most pointed answers weren’t about technology at all. They were about the decisions made around it.
Several Power readers pushed back on cloud-first mandates coming from company leaders. They said their leadership favored moving everything to the cloud on the assumption that it’s cheaper, while they believe a hybrid approach is the better bet. Others criticized senior leaders chasing the latest shiny object without weighing the cost of constant change on reliability and recoverability. A few worried about their organizations moving away from IBM i altogether.
The other recurring concern is people. IBM Z readers worried about lost COBOL knowledge and the lack of successors. “Who or what is replacing me when I retire,” one wrote, adding that it doesn’t keep them up, “but it should be keeping someone up.” Power readers framed it more personally, citing their own careers and AI eliminating entry-level and programming jobs.
Not everyone is losing sleep. Fifteen readers said nothing keeps them up, and a few joked that coffee, music, lights and owls were the culprits.
What Readers Want to Read
Readers’ content interests track closely with their day-to-day responsibilities. IBM Z readers ranked IT infrastructure and optimization (60%), systems management (55%) and AI/AIOps (53%) as their top topics. Power readers put systems management and databases first (58% each), followed by IT infrastructure and optimization (56%), and application development and security (49% each).
Articles on new technologies are the clear favorite in terms of format, chosen by 73% of Z readers and 90% of Power readers. Power readers also showed a strong appetite for practical guidance. Tutorials and technology tips were desired by 68%, with another 62% favoring webinars and 57% wanting video tutorials. Webinars were the No. 2 format for Z readers at 56%.
Readers are also looking beyond IBM. Close to 90% of Z readers and 78% of Power readers want to learn about other technologies. That reflects the environments they work in. Microsoft Windows runs in 58% of Z readers’ organizations and 77% of Power readers’, and Azure and AWS each show up in roughly 40% or more of both groups. Hybrid is the reality of their work.
TechChannel is also one voice among many. Z readers lean heavily on IBM Communities (80%) and SHARE (56%). Power readers turn to IT Jungle (70%), IBM Communities (67%), COMMON (65%) and IBM i user groups (62%). On social media, LinkedIn is the go-to platform for both groups at 73%.
How Readers Rate TechChannel
We also asked readers to grade us. Nearly all readers agree that TechChannel and our TechBeat newsletter are trusted sources of relevant news and information. More than 90% of both groups agree that we connect them to the community and the voices of the industry.
Readers act on what they read. Sharing a link to an article is the most common action, reported by 47% of IBM Z readers and 60% of Power readers. Many have forwarded the newsletter to a colleague, discussed something they read or used the information in a business meeting, proposal or program.
Earning Trust, One Issue at a Time
The survey shows a community that’s curious about what’s next and cautious about what could break. Our readers are putting AI to work, defending their systems against an ever-growing list of external and internal threats, and doing it all with lean teams, rising expectations and unclear succession planning.
That gives us a clear road map. In the year ahead, TechChannel will keep expanding practical coverage of AI on IBM Z and IBM Power, with an emphasis on governance, security and real-world use cases. We’ll continue to invest in how-to content, host webinars and provide expert voices on security, resilience and systems management. And we’ll keep looking beyond IBM to cover the hybrid environments our readers actually run.
Thank you to everyone who took the time to respond. You trust us to help you separate what matters from what doesn’t. We’ll continue to work to earn that trust.