One banking executive says the real value of AI isn’t how much it’s used—it’s what it actually delivers.

WHAT’S HAPPENING

Charles Holive, Chief AI Officer at BNP Paribas Corporate & Institutional Banking, is challenging one of Silicon Valley’s favorite AI metrics: token usage.

Rather than judging success by how many prompts employees submit or how many tokens an AI model processes, Holive evaluates AI through a business lens. He asks employees what new tasks they can accomplish, how much time they’ve saved, and whether AI is producing measurable improvements in productivity and revenue.

His teams establish clear goals, track key performance indicators (KPIs), and regularly assess whether AI initiatives are generating meaningful outcomes.

WHY IT MATTERS

As companies pour billions into AI, investors and executives are increasingly asking a simple question:

Is it actually creating value?

High adoption numbers and soaring token counts may signal activity, but they don’t necessarily prove that businesses are operating more efficiently, generating more revenue, or improving customer outcomes.

Holive’s approach reflects a broader shift from AI experimentation to AI accountability.

WHO BENEFITS

Businesses Focused On ROI — Companies measuring outcomes can better justify AI investments.

Executives And Shareholders — Clear productivity and revenue targets provide stronger evidence of value creation.

Employees Using AI Effectively — Workers who leverage AI to solve real problems can demonstrate measurable impact.

Customers — Better execution and efficiency can translate into improved service and experiences.

WHO LOSES

Vanity Metrics — Token counts and usage statistics alone may become less meaningful.

AI Initiatives Without Clear Objectives — Projects lacking measurable business goals could face increased scrutiny.

Organizations Chasing Hype — Companies focused on appearing innovative rather than delivering results may struggle to prove returns.

WHAT HAPPENS NEXT

Expect more organizations to shift from asking, “How much AI are we using?” to asking, “What business outcomes is AI producing?”

As AI adoption matures, success may increasingly be defined by revenue growth, cost savings, customer satisfaction, and productivity gains—not by the volume of prompts submitted.

The companies that thrive could be those that treat AI not as a technology experiment, but as a business strategy.

The Bottom Line: The next phase of the AI race may not be won by those generating the most tokens—it may belong to those turning AI activity into measurable business results.