The 2025 Gartner IT Symposium/Xpo in Orlando was ostensibly all about AI – generative AI, agentic AI, AI here, AI there. But talk to the CIOs and tech leaders wandering the expo halls, and a different narrative emerged. Yes, AI dominated keynotes and nearly every vendor banner, but a palpable fatigue lurked beneath the enthusiasm. One moment you’d hear, “So, did you learn anything new about AI?” By mid-week, it turned into a running joke: “Are you sick of AI yet?”
This year’s Symposium delivered both: genuine insights into enterprise technology’s future and a healthy dose of skepticism about the latest hype. In this recap, we’ll hit the highlights – from grand keynotes on finding the “Golden Path” to value, to candid hallway chatter about what’s actually working (or not) in the enterprise. Consider it a guided tour for CIOs who want the real story, with a side of pragmatism to wash down all the buzzwords.
Gartner kicked off the week by reminding everyone that while AI may be “touching everything,” it’s not everything. Daryl Plummer and Alicia Mullery urged CIOs to walk the “Golden Path to Value” — somewhere between blind AI enthusiasm and skeptical eye-rolling. Their call to action: treat AI as transformative but not magic. Real value will come from human readiness — training, governance, and a healthy respect for change management — not just deploying another chatbot. Gartner framed this as a “hero moment” for CIOs, who must turn AI’s 74% productivity gains into something CFOs can quantify (only 11% currently can).
In short: don’t chase the hype; clean your data, upskill your teams, and start laying tracks for the autonomous future before it runs you over.
Salesforce CEO Marc Benioff took the stage to evangelize his “Agentic Enterprise” vision: a world where AI agents and humans collaborate through Salesforce’s Agentforce 360. He delivered the expected optimism, but to his credit, also a dose of realism: no one, not even Salesforce, has cracked the AI ROI code. “If you think anybody has the complete answer, they’ve made a mistake,” he admitted. Benioff pointed out that technology is outpacing customers’ ability to adopt it and argued that vendors must shift from selling to partnering, with usage-based pricing and flexible models that help customers experiment safely.
The takeaway? Even the loudest AI champions know it’s time to slow down, get practical, and stop pretending every AI deployment prints money.
After two days of AI worship, Angela Ahrendts, former Burberry CEO and Apple retail executive, brought the conversation back to Earth and back to people. Her keynote on “leading with humility, humanity, and heart” reminded the room that innovation means nothing if it loses sight of empathy. “The most admired companies are those that touch lives, not just bottom lines,” she said, earning the kind of applause usually reserved for free coffee. Ahrendts urged CIOs to listen, stay present, and measure success in trust and impact as much as profit. In a week dominated by machine learning, her talk was a master class in human learning, a quiet but powerful reminder that the future of AI will still depend on distinctly human leadership.
While the keynotes provided vision, the real straight talk was happening on the show floor. By day 3, a noticeable shift in attendee attitude toward AI buzz had occurred. Early in the week, everyone was eager to swap notes on the latest AI announcements. But as the sessions wore on, an eye-rolling fatigue set in. One CIO joked that if he had a dollar for every booth with “AI” slapped on its banner, he could fund his entire IT budget for 2026.
Jacci Robinson, Anunta’s VP of GTM & Growth, observed this change in real time and even pivoted her strategy because of it. “By Wednesday afternoon, we toned down the AI talk on our booth. We realized every vendor was shouting about AI, and attendees had this glazed look like, ‘Ugh, not this again.’ So we switched to highlighting practical outcomes and even added a bit of humor.” In fact, her team nearly put up a tongue-in-cheek sign reading “Not Talking About AI Here” to attract weary passersby. “It was amazing – as soon as we led with something other than AI, people perked up. One visitor literally said: ‘Oh thank goodness, let’s talk about something else!’” Robinson noticed attendees gravitating towards discussions about clean data, integration, security, and its impact on people – “the unsexy stuff,” as she laughed – because that’s where their pain points are. Her anecdote shows how over-saturated the AI messaging was, and savvy vendors adjusted by cutting the fluff.
A particularly sharp insight came from Srinaath Arumgavel, Anunta’s Director of Product Management, who spent a good chunk of time touring competitor booths out of professional curiosity. His verdict? “There’s a disconnect between the branding and the tangible outcomes,” he said. In Arumgavel’s view, “CIOs are getting cynical. They’re asking, ‘Okay, you added an AI chatbot – does it actually solve my integration nightmare or reduce my technical debt?’ From what I saw, the answer is not convincing.” His advice to fellow attendees was to listen for outcomes and integration, not just AI buzzwords. If a vendor can’t articulate how their AI feature improves something meaningful (revenue, uptime, user experience), then all that glitters isn’t gold – it’s just glitter.
Indeed, a question hung in the air of the expo halls: Beyond the novelty, what new value is AI actually delivering that we couldn’t get before? For some domains, the answer is clear, but for many enterprise applications, it’s still murky. That realization had many attendees shifting from wide-eyed wonder to healthy skepticism by week’s end. “I’ve seen this hype cycle before,” an IT veteran told me over a mediocre pastry, “Remember ‘cloud washing’ a decade ago? Now it’s ‘AI washing.’ I’m more interested in which vendors can help integrate and operationalize these AI ideas we’ve already bought into.”
If there was one clear undercurrent at the Symposium, it was this: vendors have moved from showing off AI magic tricks to proving they can make everything actually work together. Platforms, not point tools. Openness, not lock-in. The show floor buzzed less with “revolutionary” demos and more with talk of orchestration: connecting data, tools, and processes across hybrid environments. Even security and governance discussions had an integration flavor, focusing on embedding oversight into existing systems instead of reinventing the wheel.
After two years of AI pilots and digital transformation projects, CIOs are exhausted by complexity. The new goal isn’t to buy more technology; it’s to make what they already have deliver. Or as one attendee put it, “We’ve sewn ourselves a beautiful Frankenstein — now we just need to bring it to life without it rampaging.”
Lenovo’s session captured that spirit perfectly. Skipping the hype, their speakers went straight to the fundamentals: hybrid infrastructure, clean data, and practical use of AI to boost human productivity. They acknowledged what many CIOs are privately admitting — most enterprises are still stuck in “pilot purgatory” with ROI nowhere in sight. Lenovo’s answer? Stop chasing shiny tools and focus on connecting what already works.
As one Lenovo exec put it, “If anybody believes they can do it end to end—from ideation to ROI to deployment—they will fail. You have to partner very specifically within your ecosystem to succeed.” It was a rare dose of honesty in a sea of AI slogans, and it landed well. The consensus across the event was clear: the future won’t belong to those shouting the loudest about AI, but to those who make it actually work.
After four days in the thick of the Symposium, the overarching takeaway for enterprise IT leaders was this: the AI gold rush isn’t yielding gold yet – but with the right groundwork, it could. There’s a collective pivot underway from frenzied hype to pragmatic execution. Organizations are shifting focus from chasing the next big AI headline to ensuring they have the plumbing and people in place to actually use AI effectively. Data needs cleaning; infrastructure needs scaling (hello, cloud budgets!); systems need to talk to each other; employees need upskilling and reassurance. It’s not glamorous, and it’s certainly not as newsworthy as “AI will solve world hunger by Q4,” but it’s what determines success or failure.
One could chuckle at how plain old IT fundamentals stole the show in an AI conference. As the National CIO Review team summed up, “AI success depends on mastering fundamentals, aligning investments with business value, and fostering cultures of learning and adaptability.” Indeed, the CIOs who came to Orlando looking for the next silver bullet were gently redirected: the silver bullet is that there is no silver bullet – just a lot of lead bullets (hard work) that, when fired in concert, get the job done.
There’s also a growing sense of AI maturity in the enterprise. The initial euphoria has passed; a bit of sobriety has set in. But that’s not a bad thing. In fact, it’s the phase when real innovation happens – when companies stop chasing every shiny object and start doubling down on what really matters for them. As Gartner’s Plummer put it, this is the time for CIOs to be heroes, by guiding their organizations through the valley of disillusionment to the plateau of productivity (to borrow the hype cycle terms). That means being the voice of reason, setting realistic expectations, and ensuring that each AI initiative ties to a clear business outcome.
Before chasing new tools, see where end-users say the real friction lives — and what IT can fix first. Get the End-User Gap Report featuring original research and first-party data to uncover where readiness gaps start (and how to close them). [Read the report]
