The AI reprieve is over. Now prove it.
Powered by Single Fin
Welcome to this week’s edition of CIOsurge!
This week:
Gary Flowers, CIO at Year Up United, on why he will not do AI for AI's sake and how he kills agents without ceremony
CIOs earned a reprieve on AI ROI, but the pressure is surging, not fading
Only the employees who use AI every day feel good about their job security
Let’s make this week a game-changer.
Stay sharp. Stay ahead.
No AI for AI's Sake: How One CIO Picks Projects, Rolls Out Licenses, and Kills Agents
Every leader I talk to has the same problem: a hundred AI ideas from the workforce and budget for ten. So I asked Gary Flowers, CIO at Year Up United, how he picks. His answer was old-school in the best way, and it came with the most liberating point I have heard about killing projects.
His first principle predates AI entirely. "I used to say, we're not going to do technology for technology's sake. Now I say, we're not going to do AI for AI's sake. It has to have a measurable output." Year Up United runs on a handful of core metrics, and every AI use case has to map to one, not an idea from a TikTok video but something that moves a number the organization already runs on.
To find those use cases, his transformation and technology department pairs a technical person with a business person for each area, each one knowing enough about the other's world to be dangerous. Together they pull use cases straight out of the department's metrics.
His rollout matched the discipline. With just under 600 people, "I'm not giving AI licenses to 600 people." He started with about 100, full coverage where daily use is expected, and named champions to spread what works. The mandate is still to swing: connect three times out of ten and you are in the Hall of Fame.
Then the part every leader should internalize about failure. "I don't feel bad about turning an agent off. It's not a person." Pause it, fix it, turn it back on. Compare that to unwinding a failed project staffed with ten humans.
The takeaway: admit only use cases that move a metric you already run the company on, roll licenses out to the passionate hundred before the indifferent six hundred, and kill underperforming agents without ceremony. As Gary put it, "you can't improve what you can't measure."
CIOs Earned an AI Reprieve, But The ROI Pressure Is Surging Anyway.
2026 arrived as the year AI ROI would need to get real, CIO.com reports. After years of experiments and pilots that largely failed to scale, CEOs' top priority for CIOs was demonstrable benefits from AI investments. The pressure was concrete: 71% of IT leaders said in February that they believed they had until midyear to prove AI value or face budget or job fallout, according to a Dataiku survey, and three-quarters admitted remorse over at least one major AI vendor or platform selection made in the previous 18 months. Six months later, past the midyear mark, CIOs who set their course for AI ROI are finding the destination remains elusive. There has been no spate of CIO firings, and AI spending continues to grow, but the reprieve has not eased the underlying pressure to prove value.
A reprieve is not an acquittal. The deadline passed, the firings did not come, and the temptation now is to exhale. That is exactly the wrong read. The pressure did not go away, it moved up a level, from "show me a pilot" to "show me returns," and the boards asking are more informed than they were a year ago. Gary's discipline in this edition is the antidote: map every AI dollar to a metric you already run the business on, and be honest about what is not working. The CIOs who treat the reprieve as breathing room will be having a much harder conversation at year-end than the ones who treat it as a warning.
— Zack Tembi
Only Daily AI Users Feel Good About Their Job Security
Employees and students preparing to enter the job market report real time-saving benefits from AI, but many fear its impact on their job security, according to a CNBC and SurveyMonkey survey covered by CIO Dive that polled 1,686 US workers and students in July. AI use is uneven: 30% use it daily, 34% a few times a week, and 37% never. Only the daily users feel positively about the technology's effect on their job security, while those who use it weekly or less say it makes them feel less secure. The report notes that organizations are often sending mixed messages, telling employees to use AI but not too much, to find ROI but not to overspend on tokens, leaving people to locate the Goldilocks zone of usage and value while still doing their day jobs.
This is a leadership problem wearing a workforce costume. The people who use AI daily feel secure because familiarity replaced fear with a sense of control, and the 37% who never touch it are exactly the ones most afraid of it. Mixed messaging from the top makes it worse. Gary's champion model in this edition is the fix: give the tools first to the people who are eager, let them demonstrate the wins, and let that pull the hesitant along rather than mandating adoption from above. If your people are guessing at how much AI use is too much, that is a message you have not sent clearly enough.
— Zack Tembi
💡 CIO Spotlights
Daniel Dubowski named Chief Information Security Officer at Marriott
Appointed SVP and CISO at Marriott International, reporting to Global CIO Naveen Manga, and responsible for strengthening cybersecurity and resilience across the global enterprise
Joins from Hertz, where he also served as CISO
Succeeds Arno Van Der Walt
Brings extensive experience leading cybersecurity and risk programs across the hospitality, travel, financial services, and healthcare sectors
Steps in as Marriott advances a multiyear technology transformation, including conversational search and automation testing across hotel operations
Sam Ishak named Chief Information Officer at Teranet
Appointed CIO of Teranet, Canada's leader in statutory registry services and property intelligence, leading enterprise technology strategy across architecture, application development, information security, infrastructure, and business intelligence
Brings more than 20 years leading enterprise technology strategy and transformation across financial services, insurance, and government
Previously CIO at FCT, where he led complex technology modernization initiatives
Licensed Professional Engineer in Ontario and a Distinguished Certified IT Chief Architect through The Open Group; received the 2025 Futurist Toronto CTO Award
