In this edition of Ashling Unfiltered, we put MIT Nanda's The GenAI Divide State of AI in Business 2025 under the microscope. We pull the headline claim, give you our verdict, check it against field reality, and end with one move you can make this quarter. Let's dive in.
If you read MIT Nanda’s GenAI Divide and felt whiplash, you are not alone. The report cites three different stats:
Readers are left asking the right question: which 5 percent matters?
Part of the confusion comes from the catch-all phrase "AI tools". It blurs the difference between knowledge agents that retrieve and summarize and autonomous agents that decide and act. Each type of "AI tool" or "AI agent" differs in role, function, limitations, and levels of autonomy—and those distinctions matter. Treat them as interchangeable and you risk mismatched business expectations and outcomes.
The deeper point is this: businesses that successfully cross the divide follow an effective operating model and approach. Put simply, they start small, think big, and scale quickly.
The good news is there’s a proven path that avoids common pitfalls. With the right approach, grounded in transparency, experience, and patience, results can be very different.
ROI remains critical, that will never change, but mature programs recognize that ROI takes on many forms: business leverage, payback period, capability validation, etc. These are business KPI’s beyond efficiency gains or straight cost cutting. Success also depends on consistent execution and capturing ongoing value across the program, without underestimating the role that people play. Ultimately, it’s people who bring multi-modal automation together.
The opportunity with AI—and any transformative technology—is to balance rapid experimentation with meaningful business impact. Speed in execution can still be achieved, but the most effective path begins with enhancing existing automations, where proven processes can be extended to unlock greater value.
Where those opportunities don’t exist, build and qualify your automation pipeline ensuring a steady flow of high-impact candidates. Build a portfolio of automation opportunities that not only validates emerging technologies but also drives measurable outcomes for the business.
If you’re looking to learn more about AI and Agentic processes that have been deployed to production and the realities of what it takes to get them there, bring Ashling in. Our Discovery Diagnostic is a scalable, repeatable framework that uses a structured, top-down approach to qualify, quantify, and prioritize process automation opportunities. The diagnostic begins with a process inventory grounded in how work is truly performed, enabling the identification and qualification of AI and automation candidates with clarity.