Eric Vaughan, CEO of enterprise-software powerhouse IgniteTech, was unwavering as he mirrored on essentially the most radical choice of his decades-long profession. In early 2023, satisfied generative AI was an “existential” transformation, Vaughan checked out his group and noticed a workforce not totally on board. His final response: He ripped the corporate right down to the studs, changing practically 80% of employees inside a yr, in accordance with headcount figures reviewed by Fortune.
Over the course of 2023 and into the primary quarter of 2024, Vaughan advised Fortune, IgniteTech changed lots of of workers, declining to reveal a particular quantity. “That was not our purpose,” he advised Fortune. “It was extraordinarily tough … However altering minds was tougher than including expertise.” It was, by any measure, a brutal reckoning—however Vaughan insists it was needed, and mentioned he’d do it once more.
For Vaughan, the writing on the wall was clear and dramatic.
“In early 2023, we noticed the sunshine,” he advised Fortune in an August 2025 interview, including he believed each tech firm was going through a vital inflection level round adoption of synthetic intelligence. “Now I’ve definitely morphed to consider that that is each firm, and I imply that actually each firm, is going through an existential menace by this transformation.”
The place others noticed promise, Vaughan noticed urgency—believing failing to get forward on AI might doom even essentially the most strong enterprise. He known as an all-hands assembly along with his international distant group. Gone have been the snug routines and quarterly targets. As a substitute, his message was direct: All the things would now revolve round AI. “We’re going to offer a present to every of you. And that reward is great funding of time, instruments, schooling, initiatives … to offer you a brand new talent,” he defined. The corporate started reimbursing for AI instruments and prompt-engineering lessons, and even introduced in outdoors consultants to evangelize.
“Each single Monday was known as ‘AI Monday,’” Vaughan mentioned, along with his mandate for workers that they might work solely on AI. “You couldn’t have buyer calls; you couldn’t work on budgets; you needed to solely work on AI initiatives.” He mentioned this occurred throughout the board, not only for tech employees, but additionally for gross sales, advertising, and everyone else at IgniteTech. “That tradition wanted to be constructed. That was the important thing.”
This was a significant funding, he added: 20% of payroll was devoted to a mass-learning initiative, and it failed due to mass resistance, even sabotage. Perception, Vaughan found, is a tough factor to fabricate.
“In these early days, we did get resistance, we received flat-out, ‘Yeah, I’m not going to do that’ resistance,” he mentioned. “And so we mentioned goodbye to these folks.”
The pushback: white collar resistance
Vaughan was stunned to search out it was usually the technical employees, not advertising or gross sales, who dug of their heels. They have been the “most resistant,” he mentioned, voicing numerous considerations about what the AI couldn’t do, moderately than specializing in what it might. The advertising and salespeople have been enthused by the probabilities of working with these new instruments, he added.
This friction is borne out by broader analysis. Based on the 2025 enterprise AI adoption report by Author, an agentic AI platform for enterprises, one in three employees say they’ve “actively sabotaged” their firm’s AI rollout—a quantity that jumps to 41% of millennial and Gen Z workers. This will take the type of refusing to make use of AI instruments, deliberately producing low-quality outputs, or avoiding coaching altogether. Many act out due to fears that AI will exchange their jobs, whereas others are pissed off by lackluster AI instruments or unclear technique from management.
Author’s chief technique officer Kevin Chung advised Fortune the “massive eye-opening factor” from this survey was the human factor of AI resistance.
“This sabotage isn’t as a result of they’re afraid of the know-how,” he mentioned. “It’s extra like there’s a lot strain to get it proper, after which while you’re handed one thing that doesn’t work, you get pissed off.”
He added Author’s analysis exhibits employees usually don’t belief the place their organizations are headed.
“If you’re handed one thing that isn’t fairly what you need, it’s very irritating, so the sabotage kicks in, as a result of then persons are like, ‘Okay, I’m going to run my very own factor. I’m going to go determine it out myself.’” You positively don’t need this sort of “shadow IT” in a corporation, he added.
Vaughan mentioned he didn’t need to drive anybody.
“You’ll be able to’t compel folks to alter, particularly in the event that they don’t consider,” he mentioned, including perception was actually the factor he wanted to recruit for.
Firm management in the end realized they’d need to launch an enormous recruiting effort for what turned generally known as “AI innovation specialists.” This utilized throughout the board: to gross sales, finance, advertising, and elsewhere. Vaughan mentioned this time was “actually tough” as issues inside the corporate have been “the wrong way up … We didn’t actually fairly know the place we have been or who we have been but.”
A few key hires helped, beginning with the one that turned IgniteTech’s chief AI officer, Thibault Bridel-Bertomeu. That led to a full reorganization of the corporate that Vaughan known as “considerably uncommon.” Basically, each division got here to report into the AI group, no matter area.
This centralization, Vaughan mentioned, prevented duplication of efforts and maximized data sharing—a standard wrestle in AI adoption, the place Author’s survey exhibits 71% of the C-suite at different corporations say AI functions are being created in silos and practically half report their workers have been left to “determine generative AI out on their very own.”
No ache, no acquire?
In alternate for this tough transformation, IgniteTech reaped extraordinary outcomes. By the top of 2024, the corporate had launched two patent-pending AI options, together with a platform for AI-based e mail automation (Eloquens AI), with a radically rebuilt group.
Financially, IgniteTech remained sturdy. Vaughan disclosed the corporate, which he mentioned was within the nine-figure income vary, completed 2024 at “close to 75% Ebitda”—all whereas finishing a significant acquisition, Khoros.
“You multiply folks … give folks the power to multiply themselves and do issues at a tempo,” he mentioned, touting the corporate’s skill to construct new customer-ready merchandise in as little as 4 days, an unthinkable timeline within the outdated regime. Within the months since, Vaughan advised Fortune in an early 2026 assertion, the corporate has solely saved rising its headcount, recruiting globally for AI Innovation Specialists throughout each perform, from advertising to gross sales to finance to engineering to assist.
What does Vaughan’s story say for others? On one degree, it’s a case research within the ache and payoff of radical change administration. However his ruthless method arguably addresses many challenges recognized within the Author survey: lack of technique and funding, misalignment between IT and enterprise, and the failure to interact champions who can unlock AI’s advantages.
The ‘boy who cried wolf’ drawback
To make certain, IgniteTech is way from alone in wrestling with these challenges. Joshua Wöhle is the CEO of Mindstone, a agency that gives AI upskilling providers to workforces, coaching lots of of workers month-to-month at corporations together with Lufthansa, Hyatt, and NBA groups. He not too long ago mentioned the 2 approaches described by Vaughan—upskilling and mass alternative—in an look on BBC Enterprise Right this moment.
Wöhle contrasted the latest examples of Ikea and Klarna, arguing the previous’s instance exhibits why it’s higher to “reskill” current workers. Klarna, a Swedish buy-now, pay-later agency, drew appreciable publicity for a choice to cut back members of its buyer assist employees in a pivot to AI, solely to rehire for a similar roles.
“We’re close to the purpose the place [AI is] extra clever than most individuals doing data work. However that’s exactly why augmentation beats automation,” Wöhle wrote on LinkedIn.
A consultant for Klarna advised Fortune the corporate didn’t lay off workers, however has as an alternative adopted a number of approaches to its customer support, which is managed by outsourced customer support suppliers who’re paid in accordance with the quantity of labor required. The launch of an AI customer support assistant lowered the workload by the equal of 700 full-time brokers—from roughly 3,000 to 2,300—and the third-party suppliers redeployed these 700 employees to different shoppers, in accordance with Klarna. Now that the AI customer support agent is “dealing with extra advanced queries than after we launched,” Klarna says, that quantity has fallen to 2,200. Klarna says its contractor has rehired simply two folks in a pilot program designed to mix extremely skilled human assist employees with AI to ship excellent customer support.
In an interview with Fortune, Wöhle mentioned one shopper of his has been very blunt along with his employees, ordering them to dedicate all Fridays to AI retraining, and in the event that they didn’t report again on any of their work, they have been invited to depart the corporate.
He mentioned it may be “kinder” to dismiss employees who’re proof against AI: “The tempo of change is so quick that it’s the kinder factor to drive folks via it.” He added he used to suppose if he received all employees to essentially love studying, then that might assist Mindstone make an actual distinction, however he found after coaching actually hundreds of people who “most individuals hate studying. They’d keep away from it if they’ll.”
Wöhle attributed a lot of the AI resistance within the workforce to a “boy who cried wolf” drawback from the tech sector, citing NFTs and blockchain as applied sciences that have been billed as revolutionary however “didn’t have the true impact” that tech leaders promised.
“You’ll be able to’t actually blame them” for resisting, he mentioned. Most individuals “get caught as a result of they suppose from their work circulate first,” he added, they usually conclude AI is overhyped as a result of they need AI to suit into their outdated method of working. “It takes much more pondering and much more sort of prodding so that you can change the way in which that you simply work,” however when you do, you see dramatic will increase. A human can’t presumably maintain 5 name transcripts of their head whilst you’re making an attempt to jot down a proposal to a shopper, he affords, however AI can.
Ikea echoed Wöhle when reached for remark, saying its “people-first AI method focuses on augmentation, not automation.” A spokesperson mentioned Ikea is utilizing AI to automate duties, not jobs, liberating up time for value-added, human-centric work.
The Author report notes corporations with formal AI methods are much more prone to succeed, and people who closely spend money on AI outperform their friends by a big margin. However as Vaughan’s expertise exhibits, funding with out perception and buy-in will be wasted power. “The tradition wanted to be constructed. In the end, we ended up having to exit and recruit and rent people who have been already of the identical thoughts. Altering minds was tougher than including expertise.”
From the vantage level of early 2026, Vaughan mirrored in an announcement to Fortune, month-to-month all-hands conferences look nothing like they used to: “We killed the format of reviewing targets and metrics. Now groups demo what they constructed.” He wished to emphasize one thing else: Regardless of the drastic actions he took to restructure, he nonetheless doesn’t suppose he’s forward of the curve.
“We’re simply not getting run over from behind but,” he mentioned. “The tempo of change in AI is relentless. If we don’t maintain pushing, continue learning each single day, we’re toast.”
For Vaughan, there’s no ambiguity. Would he do it once more? He doesn’t hesitate: He’d moderately endure months of ache and construct a brand new, AI-driven basis from scratch than let a corporation drift into irrelevance.
“This isn’t a tech change. It’s a cultural change, and it’s a enterprise change,” he mentioned, including he doesn’t suggest others observe his lead and swap out 80% of their employees.
“I don’t suggest that in any respect,” he mentioned. “That was not our purpose. It was extraordinarily tough.”
However on the finish of the day, he added, everyone’s received to be in the identical boat, rowing in the identical route. In any other case, “we don’t get the place we’re going.”
A model of this story was printed on Fortune.com on August 17, 2025.
