The Relocation of Value
Tiffany Treacy · AI for Good Global Summit · Geneva · July 8, 2026
AI for Good Global Summit · Geneva · July 2026
The Relocation
of Value
What AI Means for How We Lead
Tiffany Treacy
VP of Product, Power Platform · Microsoft
Presenting at
AI for Good
Global Summit
Geneva, Switzerland
July 8, 2026
Organized by ITU · United Nations
Your people are already using AI.
The question is whether your organization is ready to benefit from it.
The new agency equation.
As agents take on more execution, humans gain more room to direct the work — to make the calls, set the constraints, and own the outcomes.
The people pulling ahead aren't the ones replacing their judgment with AI. They're the ones using AI to extend what their judgment can reach.
Research scope
- Trillions of anonymized Microsoft 365 signals
- 20,000 AI users surveyed globally
- 10 markets, multiple sectors
- Validated with leading AI researchers
The finding that matters most
"Often, individuals are ready.
The systems around them are not."
Three shifts already underway
Today
Tomorrow
Isolated tools
Thinking partners
Knowledge in siloes
Shared context
Switching between apps
Intent-driven orchestration
"Often, individuals are ready. The systems around them are not."
of AI's real-world impact comes from organizational factors
Culture, manager behavior, and talent practices
from individual factors alone
Mindset, skill, and individual behavior
You are the variable.
Most organizations are stuck here.
Between starting with AI and actually changing how work gets done.
AI Adoption
"Did we start using AI?"
Seats purchased, logins provisioned, pilots kicked off
A moment in time
Tools get abandoned — shelfware by quarter three
Surface-level access
Inputs change; process, culture, outcomes don't
AI Absorption
"Has AI changed how we actually operate?"
Work redesigned, value measurable, outcomes shifted
A sustained, compounding process
Capability becomes institutional; it outlasts any one tool
Embedded in muscle memory
Part of how decisions get made, not just how tasks get done
Reimagine business processes
Build learning systems
Set culture
Intelligence is available on demand. Responsible direction is still yours.
Earlier management eras were defined by the design of scale. This era will be defined by the design of judgment, learning, and coordinated action — across humans and machines.
The organizations that pull ahead won't have more AI licenses. They'll have operating models built for a world where intelligence is abundant and human direction is the premium.
The old operating model
Linear. Step-by-step. Predictable. Slow.
The new operating model
Adaptive. Judgment-centered. Coordinated.
A learning organization isn't one that runs the most experiments.
It's one where a single person's discovery becomes everyone's capability.
more follow-up visits
when health workers could spend time on patients, not paperwork
In a rural clinic in Kenya, a program officer named Maya noticed something: health workers spent hours each week writing visit reports by hand.
She tried an AI tool that let them dictate notes in Swahili and auto-generate the report. Workers spent that time on patients instead of paperwork — and follow-up visits jumped 30%.
Without a learning system — that's where the story ends. A nice local win. A slide in one team's update. Forgotten in a quarter.
With a learning system — here's what happens instead.
A continuous loop, not a one-time pilot
01 · Capture
Log every experiment
Problem, approach, result — in one shared place. Owned by no single team.
02 · Codify
Turn it into an asset
A vetted template, guide, and privacy checklist others can actually use.
03 · Circulate
Push to every team
Facing the same problem — source named, so peers can ask follow-up.
04 · Compound
Each improvement flows back
The asset gets smarter with every use. The loop closes and restarts.
Maya's 30% becomes a capability the whole organization owns — and it keeps improving without central command.
The biggest thing you can do this week isn't launch a new tool.
It's be visible about how you use the ones you have.
points
AI value when managers openly model its use
points
critical thinking about when and how to use AI
points
trust in AI agents doing complex multi-step tasks
more likely to be high-frequency AI users when managers create space to experiment
Here's what absorption looks like at humanitarian scale.
"That's not adoption. That's absorption."
The IOM needed to know where climate risk was concentrated — before a crisis, not after. Microsoft brought AI expertise and compute capacity. IOM brought mission and local ground truth from partners in Libya, Ethiopia, and the Maldives.
Together, they mapped vulnerable communities at a precision that hadn't existed before — identifying populations at risk of extreme heat and flooding, proactively, so resources could move first.
All three leadership responsibilities — in one program
- Reimagined how risk identification actually works
- Built a learning system across three countries and multiple partners
- Set a culture of shared ground truth over institutional silos
AI isn't just making existing workers more productive.
It's inviting people who couldn't fully participate before.
A Microsoft study on AI and disability found that for people who have spent years navigating systems not built for them, AI tools are reducing barriers that existed for decades — in communication, navigation, information access, and work itself.
"This is the one I never hear people talk about. And every time I raise it, the room changes."
The organizations that pull ahead won't run the most experiments.
They'll learn the fastest across all of them.
Name one process you'll reimagine — not tweak.
Start from your mission, not from habit.
Name one local win that should become org-wide capability this quarter.
That's the seed of your learning system.
Model it yourself. Your team is watching.
Leadership is the variable. The relocation of value is already underway.
Tiffany Treacy · VP of Product, Power Platform · Microsoft · AI for Good Global Summit · Geneva · July 8, 2026