Special edition · Signal4i reads the field

Where the community stands.

Adoption is nearly everywhere. Readiness is not. That gap runs through all of 2026, and it's why a community measuring itself now is worth two minutes.

Two years ago the question was whether AI would matter. That question is closed. The 2026 surveys agree on the top-line to a degree that's almost boring: AI is inside nearly every enterprise. What they also agree on is the part that matters. Being inside is not the same as being ready.

The numbers land in the same place from a dozen directions. Publicis Sapient's global survey of 1,550 decision-makers found that 73% use AI regularly or across most processes, but only 10% call it core to how the business runs. The report is blunt about that 63-point gap: it isn't a technology problem, it's an organizational one. Deloitte's read of over 3,000 leaders lands the same way, with most enterprises using AI and only about a quarter having scaled it. The sharpest edge comes from IDC.

88% never ship
Nearly nine in ten AI pilots never reach production. They stall on governance, data-readiness, and observability, not on the model.
IDC, 2026 · via industry reporting

The gap has a shape, and 2026 named it: the shift from generative AI to agentic AI, systems that don't just answer but act, taking multi-step actions with limited human involvement. That shift is what turns readiness from an academic question into an urgent one. A tool that suggests can be wrong quietly. An agent that acts is wrong out loud. And the field is moving fast: Schellman's 2026 governance report found 86% of organizations have already tested or piloted AI agents.

Which is exactly where the confidence and the reality come apart. That same report found 74% of enterprises believe they could pass an AI governance audit today, and only 27% actually could. Ninety percent have funded governance; the funding hasn't yet bought the thing it was for. Deloitte's version: about one in five organizations has a mature model for governing autonomous agents, even as agentic use climbs sharply. Capital is moving into deployment faster than governance is being built, which is historically the pattern that raises incident probability before it lowers it.

The field isn't short on ambition or on spend. It's short on the readiness that turns either one into an outcome you can trust.

You don't have to take that framing from a market observer. IBM's own chairman and CEO, Arvind Krishna, put it plainly from the Think 2026 keynote stage: the enterprises pulling ahead are not deploying more AI — they're redesigning how their business operates. IBM's CEO study around the same event found only 25% of enterprise AI initiatives delivering the ROI expected of them, and just 16% scaled enterprise-wide. When the vendor selling the transformation tells you the constraint is your operating model and not your technology, it's worth listening.

The IBM i community sits in a specific place in this picture

Here the platform's peculiar position becomes an advantage worth understanding rather than a liability to apologize for. The readiness the wider field is scrambling to bolt on — governed access, an audit trail, identity that means something, reversibility — is, on IBM i, largely native. That isn't a vendor claim. At COMMON's own PowerUp conference in April 2026, IBM's Adam Shedivy described IBM i as "agentic-native": the operating system, the Db2 database, the object model, user profiles, security, and the audit journals all belong to one integrated platform, so an authorized agent can observe, understand, and act inside established, governed interfaces rather than across a loose collection of disconnected tools. The things an agent needs in order to be trusted, the platform already enforces.

The counterweight is just as real, and the community knows it. The expertise that runs these systems sits with a generation moving toward retirement, and the pipeline behind it is thin. The entry-level cohort across the economy is running well below its historical hiring trajectory, and IBM i feels that more acutely than most. The platform is strong on what the field lacks and exposed on what it can't easily replace, which is people. IBM's own 2026 framing says as much: AI is being treated as a technology issue when it's increasingly a workforce one, and buying the platform is the easy part.

So the IBM i community holds an unusual hand. Strong where the field is weak, exposed where the field is only starting to look. And it holds that hand in the middle of a platform moment: IBM reports 70% of IBM i shops plan to upgrade hardware or software this year, agentic-ops capabilities are arriving, modernization tooling is maturing. A moment like that rewards a community that can see itself clearly and move together, and it punishes one that can't.

Which is why measuring the community, now, is not a formality

You can't move together on a picture you don't have. Every one of the surveys above measured the enterprise at large: global, cross-platform, cross-industry. None of them measured this community. And a community whose position is this distinct, this strong in some places and this exposed in others, is badly served by a general-industry average that washes out what makes it different.

This is where COMMON's role matters. Since 1960 it has served the IBM i community through education, certification, networking, and the word that matters most here, advocacy. That work is concrete, not abstract: this October it runs NAViGATE, its IBM i education conference, co-located with IBM TechXchange in Atlanta. COMMON represents this community's interests to IBM and to the wider industry, and its stated purpose is to enhance the careers of the people on the platform. But advocacy and education only work when they're accurate. You cannot represent a community to IBM, or decide what to teach it, against a picture you're guessing at. The organization whose job is to speak for the IBM i world needs to know where that world actually stands.

That is what the COMMON and Fortra survey is for. It reads the community's real position, from where the data lives to how ready the skills are to how governance is holding up, and turns that into the picture COMMON uses to decide what the community needs and how to advocate for it. The survey isn't COMMON asking for something. It's COMMON building the instrument it needs to serve you well.

You've just read where the field stands. The one thing this essay can't tell you is where you stand in it, and neither can COMMON, until you say so.

Eleven questions, two minutes. It's how the community that represents you finds out what it's representing.

Be counted in the community's read of itself

The State of AI in the IBM i Community

COMMON and Fortra are running a short survey on where AI stands across the IBM i world. Your answers become part of the picture COMMON uses to serve and advocate for the community, which is the difference between direction set on real data and direction set on assumption. Eleven questions, about two minutes. You've just read the whole landscape; adding your own shop to it is the quick part.

Take the community survey →
Runs on SurveyMonkey · results serve COMMON's read of the community
Signal4i reads the field · a special edition, outside the weekly Orientation cadence.
The survey is COMMON + Fortra's, hosted on SurveyMonkey. This page states where things stand and points to the community's own measurement.
Sources: Publicis Sapient 2026 Global Enterprise AI Report · Deloitte 2026 State of AI in the Enterprise · Schellman 2026 State of AI Governance · IDC via industry reporting · Arvind Krishna / IBM CEO study, IBM Think 2026 keynote (May 2026) · Seiden Group / IBM (Adam Shedivy, COMMON PowerUp, Apr 2026) · IT Jungle (IBM i Marketplace Survey) · IBM Think 2026. Figures cited from published 2026 reporting.