The baseline for Fortune 500 AI transformation. An index built from evidence, not hype.
.png)
.png)
Every month, the Barometer reads the evidence for what the Fortune 500 is running:
The Barometer scores five measures, read in the order transformation actually happens:
It reads three groups against them - the broad Fortune 500, tech and startups as the pace-setter, and a drill-in on financial services - so a leadership team sees where its segment sits and what's holding it back.
Every month, the Barometer turns the data into two things a leader can use:
This is the Momentum Factor: whether a segment's AI spending is running ahead of its ability to govern, adopt, and prove it.
The Barometer follows two numbers - what a segment spends on AI, and what it reports back in earnings.
We get this question a lot. “Agentic” describes AI that does more than answer - it decides and it acts. We define AI agency by its level of autonomy and decision-making ability, and the samples below are based on what we’re seeing in the marketplace:
.svg){kind=icon}
This can be some general info about "assistive"
What it does
Produces something when asked: a draft, a summary, a classification, a suggested reply. It does not act on its own output.
Who decides
A person, every time. Nothing reaches a customer or another system unless someone puts it there.
Typical use cases
Summarizing internal documents and answering questions from them, suggesting an email reply for a person to send, extracting and classifying data, code assistance.
This can be some general info about "delegated"
What it does
Takes an instruction and completes the steps itself, pulling data and calling other software without being walked through each decision.
Who decides
The system chooses sequence and method inside limits set in advance. A person sets the boundary, not each action.
Typical use cases
Account intake and opening, portfolio rebalancing inside a scoped account, scheduling and booking appointments, shipment and order updates, refund handling.
This can be some general info about "orchestrated"
What it does
Several agents run a process end to end, handing work between them, with one directing the others and auditing the output.
Who decides
Mostly the system. A person moves to design, monitoring and handling exceptions.
Typical use cases
Autonomous procure-to-pay, supply chains that reorder and reschedule themselves, period-end close, contract lifecycle handling.
Most of what your leadership team believes about enterprise AI readiness came from someone selling it. The Agentic Barometer™ is the monthly correction. It reads how far the Fortune 500 has really moved on AI transformation and answers the two questions a board keeps asking: where the enterprise actually stands, and how fast it should be moving.
First principles to enterprise impact™
