This post is authored by Ross Leibowitz, a seasoned expert with over 25 years of experience in real estate and workplace solutions. As a founder of Centerstone Software and former CTO of Manhattan Software, Ross has spent his career building and innovating workplace products. As Tango’s Senior Director of Workplace Products, he continues to drive product strategy and shares his unique perspective on how AI is reshaping the modern workplace.
In my first blog, we laid out the inevitability of disruption caused by AI. This one is about the next step—the human one. Here’s the truth: you can buy all the best AI tools, but if your teams are afraid of or uncomfortable with them, adoption will stall before it even starts.
In this post, I’ll explain why I believe the most important step in the next phase of AI isn’t engineering or licensing… It’s cultural. It’s change management. We need to build a path from awareness to trust, from fear to use. Only then can we unlock real value in practical applications.
From disruption to discovery: Why getting comfortable matters
In my last piece, “Preparing for Disruption,” the focus was on getting people to accept that AI is coming, whether they like it or not. But acceptance is a low bar. What we really care about is use. And use requires a certain level of comfort.
If someone doesn’t trust AI, or they think it’s risky, they won’t turn to it or even work against it. They’ll go back to what’s familiar: spreadsheets, long email chains, manual processes.
In my experience working with workplace and facilities leaders, the ones that succeed aren’t those with the flashiest tools, they’re the ones who get their people just comfortable enough to start doing small tasks with AI. Over time, confidence grows, curiosity spreads, and the tools stop being scary.
Data backs this up. A recent BCG report notes that while AI adoption is growing, only 13% of companies have deeply integrated AI agents into workflows. Many are stuck in pilots or experimental phases. Frontline adoption has stagnated.
In short: adoption is not zero-sum, but it is fragile. The difference between “we tried AI” and “we use AI” is trust, comfort, and consistent practice.
The fear axis: What I heard at IFMA
At IFMA’s World Workplace this year, I heard a lot of the same sentiment: AI makes me nervous. Not because they think it will take their jobs, but because they didn’t trust it.
That hesitation is real. According to PwC’s 2025 AI Predictions report, trust—not cost or capability—is now the top barrier to adoption. The Edelman Trust Barometer found that over half of employees want clearer guidance before they feel comfortable using AI at work.
Mostly what I’ve heard isn’t fear of replacement, it’s fear of reliance. People are hesitant to depend on something they don’t trust or comprehend. That’s why comfort matters. Only when people feel safe experimenting will they start to see what AI can actually do for them.
So, what do we do about fear? We can’t eliminate it, but we can manage it.
Change management, not just tech management
Here’s something I say a lot: deploying AI is a change initiative dressed up as a tech initiative.
If you treat it like any other software rollout, it will fail.
The real work isn’t about tools. It’s about people—habits, mindsets, expectations. The tactics we use for any large organizational change apply, but with nuance.
So how do you build that foundation?
Psychological safety is nonnegotiable. Your teams need a “sandbox” where they can experiment, make mistakes and learn, without retribution. This will help build confidence.
Guided cohorts can help. For example, a pilot group or “AI buddies” cohort can try tools together, share learnings, and help each other over the hump.
Visible leadership matters. If executives or department heads show they’re experimenting with AI personally, not just mandating it, it changes the narrative from “this is risky” to “we’re all learning.”
Feedback loops are essential. Set dedicated times for discussing what’s working (or not). Over time, those iterations build trust and build safe paths for more people to join in.
Here is a compelling supporting point. The MIT report on cultural benefits of AI found that in organizations where AI improved efficiency and decisions, over 75% also saw upticks in team morale, collaboration, and collective learning. Another BCG study observed that 58% of companies saw increased efficiency and decision quality after AI deployments—and of those, nearly all saw culture-related improvements in at least one dimension (morale, role clarity, collaboration).
Put frankly: when AI adoption is done right, culture doesn’t suffer, it has the opportunity to improve.
Designing “safe zones” for AI exploration
This is where the rubber meets the road. You have to create controlled environments where people can gain comfort before you expect competence.
Here’s how I recommend doing it:
- Low-stakes tasks first.
Remove company data from the picture and start with personal queries. For example: What to make for dinner with your existing pantry items, what television show to watch after you just got done with one you loved.
- No proprietary or sensitive data.
Use sanitized or dummy data initially, or allow teams to test in offline or sandbox environments. The goal is to remove fear of “oops, I exposed something sensitive.”
- Buddy or peer learning.
Pair less experienced users with “AI explorers” or early adopters. Let them test side-by-side or review prompts together.
- Narrative + debriefing.
After experiments, share stories: what succeeded, what failed, and what surprises emerged. Encourage vulnerability—the “AI didn’t do that” moments can be as valuable as the wins.
- Guardrails & guidelines.
Early on, provide lightweight policies about what’s okay: e.g., “Don’t ask it to generate legal contracts,” or “Don’t feed personal identifiers.” As confidence builds, expand guidelines.
In my own experience, the first time someone uses AI isn’t transformative, it just needs to make life 5–10% easier. That unlocks the curiosity to push further.
A “comfort roadmap” for your teams
To make this real, here’s a simple roadmap you can follow (and adapt). Move through it at your own pace. The goal isn’t speed, it’s confidence.
| Phase | Objective | Actions |
| Phase 1 – Gentle Introduction | Get all hands to try something basic | Assign a simple AI task (e.g. prompt for meeting notes, draft an email) |
| Phase 2 – Share & Normalize | Build collective narrative and reduce fear | Host a demo session or “show & tell” of interesting prompts |
| Phase 3 – Guided Experiments | Deepen usage in safe zones | Run small team pilots (e.g. AI in facilities scheduling, space planning) |
| Phase 4 – Guardrail Adjustment | Expand scope with oversight | Update usage policies, broaden data access, start integrating AI into workflows |
| Phase 5 – Pilot → Adoption | Turn comfort into competence | Scale the pilots into production workflows; monitor adoption and impact |
There’s no need to rush through these phases. They should welcome and incorporate feedback from the team. The goal here is actual adoption, not just more hype about AI.
Why comfort is the catalyst
I’ll leave you with a final statement: comfort before competence; trust before scale.
You can’t force your organization into AI, they have to step into it willingly. When they feel safe enough to experience, curiosity takes over. As they experiment, ask questions, fail, and retry, they internalize what works and begin to see AI as a co-worker, not a threat.
When that shift begins, the hard part (the technology, integration, scale) still remains. But you’ve already done the foundational work. You’ve primed the soil.
So, my challenge to you: over the next 8 weeks, don’t push for fully functional AI workflows. Push for experiments. Push for stories. Push for small wins. Let your teams dip in, get curious, and build confidence. Because when comfort takes hold, adoption follows.
In the next post in this series, we’ll move from mindset to maestros, showing how real AI applications are shaping workplace strategy in CRE, facilities, and FM settings. But only once comfort has been established can those tools land for real.
